AND OF THE MURRAY RIVER REGION RENMARK, AUSTRALIA · GEOLOGY AND GEOMORPHOLOGY OF THE MURRAYRIVER...
Transcript of AND OF THE MURRAY RIVER REGION RENMARK, AUSTRALIA · GEOLOGY AND GEOMORPHOLOGY OF THE MURRAYRIVER...
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVERREGION BETWEEN MILDURA AND RENMARK, AUSTRALIA
By Edmund D. Gill
Deputy Director, National Museum of Victoria, and Director of Chowilla Research Project
Introduction
"To a person uninstructed in natural history,
his country or seaside is a walk through a gal-
lery filled with wonderful works of art, nine-
tenths of which have their faces turned to the
wall". So said T. H. Huxley. Although in-
structed in natural history, we found the area
studied in this Memoir so dry in its time of deepdrought when we first went there, so flat, andso unvaried compared with other research areas,
that we wondered what story it had to yield.
However, it was judged that this country, like
all others scientists have investigated, wouldhave a useful and fascinating story when it hadbeen deciphered. So it proved to be.
The Research Project
The flattest continent in the world is Aus-tralia, and it is also the driest. Put these two
facts together, and it can be inferred that the
major river system is somewhat unusual. Suchthe Murray/Darling system certainly is, as the
sequel clearly shows. The river tract between
Mildura and Renmark is a significant part of
this system. Here the Darling River and the
Darling Anabranch enter the Murray. Herethe valley is 15 to 30 km wide, then suddenly
narrows to five kilometers, so making the pro-
posed Chowilla Dam possible. So flat is this
area that, if the dam is completed, 1370 km2
will be inundated. Lake Victoria, a giant billa-
bong (oxbow) system, is on an anabranch
formed by the Frenchman's Creek and the
Rufus River. It has acted as a massive sand
trap, so that a large dune tract lies on its E.
bank, emplaced and remodelled over 20,000
years at least by the persistent W. winds of that
country. Under natural conditions Lake Vic-
toria is half dry (except in flood time) being
an abandoned part of the course of the Murray.
At present it is used as a water storage (1 1,200
hectares).
This Research Project was triggered by the
commencement of building operations at the
Chowilla Dam site. The Trustees thought that
this little-known area where three States meet
should be investigated before it was inundated.
The undertaking was therefore a salvage one,
but on a scale not attempted in Australia be-
fore. The area to be flooded plus a necessary
marginal area gave a total of 2600 km2 to be
studied. As a result, the study had to be of a
reconnaissance nature, with more detailed at-
tention to significant sites. No account of the
geomorphology, geology and archaeology of
the area has been given before, but there are
numerous incidental references in a wide
literature; an attempt has been made to provide
a bibliography of the more important refer-
ences. The stated aim of the Project was to
collect data and materials that would have been
lost as a result of the construction of the dam,
but the ultimate aim was to achieve a funda-
mental understanding of this tract of country
—
how it came into existence and what the pre-
sent processes are. A contribution to this aim
is set out in this Memoir.
The Project was successful beyond expecta-
tion. The fundamental geology is of wide ap-
plication. The palaeontology is making import-
ant contributions to knowledge, e.g. Late
Pliocene (?) fish remains include Neoceratodus,
and the skeletons of extinct marsupials makepossible the indentification of post-skeletal
material previously not referable to any species.
Some skeletons were still articulated. Thearchaeology reveals middens back to 18,000
years ago. The Aborigines of the area were very
conservative and kept to their old core/flake
https://doi.org/10.24199/j.mmv.1973.34.01 9 May 1973
KDMUND D. GILL
culture, eschewing the new blade culture with
hafted tools. An Aboriginal skeleton with
gypsum widow's cap proves use of this device
back to c. 750 years at least. The palaeoclima-
tology permits recognition of the beginning of
the present aridity.
This extensive research operation would nothave been possible without supporting funds.
The William Buckland Foundation made the
chief contributions, and the Sunshine Founda-tion also contributed. E. D. Gill received a
grant from the Nuffield Foundation towards the
geological research. Sir Robert Blackwood,Chairman of Trustees (now President of the
Council under a new Act), took a special in-
terest in the Project, and joined most of thefield expeditions; he has presented in this
Memoir his work with K. N. G. Simpson on the
excavation of Aboriginal skeletons. The fundsprovided by the Foundations made it possible
to employ Mr. K. N. G. Simpson as Field Of-ficer. We are grateful to the scientists whohave contributed papers to this Memoir, and to
those whose assistance is acknowledged in the
appropriate places. Mr. G. Douglas of Werrimulgave us a great deal of assistance in the field.
We are much indebted to the people on theland. Their hospitality, their assistance whenneeded, and their fund of local informationgreatly assisted us. Through their help we wereable to achieve much more than would other-wise have been possible.
An account of this investigation was given in
ten radio lectures requested by the AustralianBroadcasting Commission. These were later
published under the title "Rivers of History"(Gill 1970).
Major river system of the flattest continent
Our planet has six continents of which twoare islands—Australia and Antarctica. Inaverage elevation above the sea, Antarctica is
the highest and Australia the lowest. More thanhalf of Australia is below 300 m, and only aboutfive per cent is above 600 m. Australia is theflattest continent in the world, which fact hasa profound effect on its major river-system,
the Murray/Darling. The names of the tworivers are commonly linked in this manner be-
cause the branch river (Darling) is 160 kmlonger than the main river (Murray), which is
2570 km long.
Moreover, Australia is the world's driest con-
tinent with 75 per cent of its land surface arid
or semi-arid. That Australia is so fiat and so
dry, explains many of the unusual features of
the Murray/Darling. Both rivers rise in high
rainfall areas of the Great Dividing Range.The Murray rises in the temperate zone on the
Kosciusko Plateau. The Darling rises in the
subtropical to tropical areas further north.
After only 300 km or so these rivers flow
with a very low grade (because the continent
is so flat) through 1500-2500 km of river
course which contributes little or no water (be-
cause the continent is so dry). Along the courseof the Darling River from the Queenslandborder to where it joins the Murray River at
Wentworth is about 2170 km over which it
falls only 120 m. Thus over a long distance the
remarkably low mean declivity of 5 6 cm perkm is maintained.
The Murray River at Albury has a fall of
14 2 cm per km which is reduced to theminimum of about 1 6 cm per km for the last
160 km. The water at Albury takes a monthto reach the sea. That in the headwaters of theDarling takes two months or more to reach its
conjunction with the Murray at Wentworth.Water from the Dartmouth dam being con-structed on the Mitta Mitta River will takesix to eight weeks to reach S. Australia. TheMurray/Darling system has an average annualrun-off of only 15,000 m3 per km2
. TheYangtze-Kiang River in China has a run-off of
1 19 x 10G m3 per km2 per annum.In the area under study between Mildura
(Victoria) and Renmark (S. Australia) thevalley of the Murray River (PL 1) widens to
32 km, including Lake Victoria which is aformer meander system, then narrows to 4 8km at the Chowilla Dam site. Further down-stream it narrows to 16 km. It is a significantpart of the course because here the Darling andits Anabranch join the Murray, the effects oftectonics on the river course can be studied,the cliffs reveal the essential stratigraphy, manyfossil localities have been discovered, and Lake
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER
Arepunog
Fig. 1—Locality map of study area.
EDMUND D. GILL
Victoria with its large lunette on the E. side
constitutes a phenomenon of unusual interest.
For much of its course the substrate of the
river systems is clay (Browne 1934). If the
Darling and the Murray ran into deep sandy
country they could well be swallowed up, be-
cause the volume of water carried is relatively
small, and the evaporation is high (^1-8 mp.a.). The Darling travels in what is virtually
an aqueduct of clay which prevents loss bysoakage. Some irrigation channels in sandy
areas lose 80-90 per cent by soakage, which
gives an idea of what could happen to the
river waters if they had a similar substrate. In
the area of study the river flows chiefly in clay
or clayey sand that acts as an aquiclude. TheChowilla Sand is clean and can act as an
aquifer, but it is a channel sand, and so is
limited in capacity. I have seen no springs
emerging from it. Probably the Blanchetown
Clay and other clayey formations prevent water
reaching it from the surface, while the complex
interdigitation of formations limits the distance
water can travel in the channel sands.
Apart from the meanders, the river course
directions are curious and call for explanation.
To understand this part of river-history, it is
necessary to consider the geology of the
area. Apart from their courses in the mountains
and foothills, the Murray and Darling Rivers
run over a deep sedimentary basin—the MurrayBasin. This will now be considered, because
the geological history is such that what happensat depth influences what happens at the surface.
Drainage basins
In the E. half of Australia there are two large
drainage basins—the Lake Eyre and the
Murray. The Lake Eyre drainage basin is the
largest in Australia, and is large by worldstandards, being some 1,300,000 km2 in ex-
tent. Although so large a basin, it drains very
little water. Because the continent is so flat, the
declivities are low, and there are no high moun-tains round the edge of the basin. Because Aus-tralia is so dry, very little rain falls (the isohyets
range from 12 to 25 cm only) and having flat
sandy terrains to cross with a very high evapora-
tion, the water seldom reaches Lake Eyre.
Only once in living memory (Mawson 1950,
Bonython 1955) has Lake Eyre been filled.
Lake Eyre occupies a very shallow depression
a little below sealevel. What would be a small
or negligible flexure on most terrains, here
created a huge drainage basin because if the
flatness of the country.
To the SE. is the smaller Murray Basin
(Fenner 1934), which can be defined ap-
proximately by the 180 m contour. A wide
northern part is occupied by the Darling River
and its tributaries. The Darling and Paroo
Rivers meet at Wilcannia and pass through a
comparatively narrow zone which has been
named the Cobar Neck, being W. of that city.
The basin then widens again to comprehend the
Murray River and its tributaries along with the
Darling River from the N. This is often referred
to as the Murray/Darling riverine plain. Thearea discussed in this Memoir is more or less in
the middle of this plain.
The Murray and Darling begin in the Great
Dividing Range with a strongly dendritic
stream pattern (Taylor 1914). Because the in-
land is semi-arid, the tributaries rapidly dis-
appear. In the N. the Darling is the only con-
tinually flowing stream. In the S., no tributary
of consequence enters the Murray after the
Darling joins it at Wentworth. The Murrayand Darling Rivers may thus be considered
gutters that drain the waters of the W. side
of the Dividing Range across the semi-arid
inland to reach the sea to the SW.
Sedimentary basins
The surficial drainage areas of the present
are directly related to the sedimentary basins of
the past. In E. Australia there are two im-portant sedimentary basins—the Great Arte-
sian Basin to the N. and the Murray Basin in
the S. There is a narrow connecting area W. of
Cobar as there is in the present drainage basins.
It is called the Darling Corridor by Devine andPower (1970). While the Murray Basin is aunit, the Great Artesian Basin has a number of
sub-basins such as the Carpentaria Basin on the
Gulf of that name, and the Surat Basin in E.Queensland which has been a source of oil.
The Great Artesian Basin is chiefly of Mesozoic
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER
100 100 200 Kilometre*
Fig. 2—The structure of the Murray Basin (after Hills) with some additional information.
EDMUND D. GILL
strata, while the Murray Basin consists of
Mesozoic and Cainozoic strata. Although the
Murray Basin appears small beside the GreatArtesian Basin, it is nevertheless quite exten-
sive, being about twice the area of France.
Murray framed sedimentary basin
In outcrop, the Cainozoic beds of the
Murray Basin present a rounded structure, as
may be seen in the 1960 Tectonic Map of Aus-tralia published by the Bureau of Mineral Re-sources. However, the fundamental bedrockstructure is quadrate, as shown by Hills (1956)whose map is reproduced as Fig. 2 with the
addition of some later information.
The Australian Oil and Gas Corporation
Limited sank bores at Renmark, Lake Victoria,
and Wentworth. With their permission, Fig. 3
is published containing the information obtained
therefrom. The palaeontologist who reported
for the Corporation was able to identify all
strata with some certainty as high as the Book-purnong Beds. These are probably Chelten-
hamian (Uppermost Miocene) and those re-
ported in the Lake Victoria bore are no doubtthe same horizon as that in a deep well at
Tareena Station (Tate 1899) on the N. side
of the Murray River between the S. Australian
border and Lake Victoria (now part of Kul-curna Station). Tate was uncertain of the age of
the occurrence, but compared the fauna with
A.O.G. NTH. RENMARK No. 1 A.O.G. WENTWORTH No. 1
25 20 15 10 5
LMILES
50 KILOMETRES
Fig. 3—Geological section between Renmark and Wentworth by courtesyof Australian Oil and Gas, with some adaptations.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER
that at Beaumaris, Victoria, which is the type
locality for the Cheltenhamian Stage. Mr. T.
A. Darragh, Curator of Fossils of this Museum,considers that the fauna is probably Chelten-
hamian, pointing out that Pelicaria coronata is
one of the characteristic species obtained fromthe well. At both Beaumaris and TareenaStation, the sediments are nearshore marine in
ecology.
At the Chowilla Dam site, Loxton Sands(Firman 1966) have been proved, and it maybe that at Tareena Station and Lake Victoria
they lie between the Cheltenhamian marinebed and the Parilla Sand (which outcrops
along the Murray River and up Salt Creek),
but this has not been established. Presumably,
the calcareous marine bed was at the base of
the Tareena well, while the Loxton Sands or
equivalent provided the aquifer (if any). Cer-
tainly the clayey Parilla Sand at or near the
surface would act as an aquiclude.
The stratigraphy revealed by the bores of
Fig. 3 shows:
1
.
A succession of alternating marine and
non-marine phases.
2. A massive fault in the basement rocks, with
a throw of the order of 600 m demon-strated, but the exact amount of dislocation
unknown. The A. O. G. geophysicists
traced this structure for some 24 km NE.of the Lake Victoria well site. The fault
is a major one, and so is here named TheLake Victoria Fault. Its orientation is that
of the Bouger anomaly contours (Bureau of
Mineral Resources, Geology and Geo-physics plans G. 239/1/1 and 2/1).
3. On the interpretation given in Fig. 3, move-ment began in the Aptian. As the undiffer-
entiated Lower Cretaceous is of the samethickness on both sides of the fault, it is
concluded that displacement occurred after
its deposition.
4. The earth movements have been slow. Kulp
(1961) gives the beginning of the Cre-
taceous as 135 m.y. and the beginning of
the Albian as 120 m.y. It has thus taken in
excess of 120 m.y. to build this part of
the Murray sedimentary basin.
5. The rate of movement has varied with time.
As the relative thicknesses of the formations
indicate, movements were more rapid in the
Aptian-Eocene and in the post-Miocene.
These belong to the Bass Strait (Gill 1964,
p. 347) and the Kosciusko Epochs (An-drews 1910) respectively. As registered bythe movements in this area, the Bass Strait
Epoch lasted for some 70 m.y., while the
Kosciusko Epoch has existed for only about
15 m.y. However, the latter epoch maybe considered as still in progress. The pre-
sent may be a lull in a longer period of
movements. The gap between the BassStrait and Kosciusko epochs in the MurrayBasin is about 45 m.y.
Basin Frame
The frame of older rocks round the basin
consists chiefly of the Adelaide Geosyncline
formations on the W. and the Tasman Geosyn-cline formations on the E. The former consists
essentially of Precambrian to Cambrian rocks,
while the latter consists chiefly of Lower andMiddle Palaeozoic rocks. Their structural re-
lationships are shown on the 1971 TectonicMap of Australia. These geosynclines providealso the frame rocks to the N. To the SW. is
the opening by which the sea reached the
Murray Basin, but this is complicated by a
horst, the Pinnaroo Block. Along the Padtha-way Ridge are highs of granitic rocks that
formed an archipelago in the Tertiary sea. Inother words, the horst formed a sill over whichthe sea passed into the basin.
Tectonics
The Cainozoic tectonics of Australia are
weak, contrasting with the powerful movementsthat characterize New Guinea to the N. andNew Zealand to the E. The rates of movementcontrast strongly. Thus in the middle of the
Murray Basin depression (Fig. 3) the averagerate of sinking since the beginning of the
Cretaceous has been ^4 m per 120,000 y.
whereas average rates calculated for various
lengths of time back into the Cainozoic are in
New Zealand ^-4 m per 100 y, and in New
EDMUND D. GILL
Guinea from 1 m in 5,000 y to 1 m in 88 y(Gill 1972). During the Cainozoic, the tectonic
movements were in the form of broad warps
that deepened basins and flexed up marginal
low-mountain areas such as the Mt. Lofty
Mts. on the W. of the Murray Basin, and the
Great Dividing Range on the E. and S.
1. Constancy of Tectonic Pattern
As a result of the mildness of the movements,
no change in the tectonic pattern of the study
area has occurred since the commencement of
the Cretaceous Period. Moreover, ancient faults
of Palaeozoic, or greater age (such as the LakeVictoria Fault) have continued to make their
influence felt (cf. Firman 1970), because the
small movements on these lineaments are
significant on a surface where slopes are so
slight. The basin is completely filled with
sediments and so the surface is about as flat
as it can be. The only steep slopes are the banks
of the river tract, which has been incised .—36m (PL 2).
2. Geomorphic Control by Deep-seated Linea-
ments
These palimpsest tectonics, as they mightjustifiably be named, are critical to the under-
standing of this area. The description of the
formation of Lake Victoria later in this con-
tribution illustrates this well. The River Murrayin the area of study flows through a terrain
of E-W dunes, but those to the N. appear to be
an older system than those to the S., so sometectonic boundary may be involved. N. of the
river, the Blanchetown Clay typically outcrops
at or near the surface, so that water tanks canbe excavated in it. The sand supply is therefore
limited. To the S., the country is more sandy,
and the irrigation channels in some areas lose
most of their water by soakage.
Hills (1956a) states that "tectonic signific-
ance must be attached to every stretch of die
River Murray, and that from trends alone aclear indication of structure—the nature re-
quiring to be investigated in each case—is
provided". Johns and Lawrence (1964) with
reference to that part of Murray Basin in NW.Victoria state, "It appears that movement along
ancient fines of weakness in two or three direc-
tional trends (approximately NW., NE. and
N.) within the basement rocks, have been
dominant in the formation of the basin". Hills
(1956b) has called this "resurgent tectonics".
3. Strong Influence of Small Faults in Flat
Country
Displacement in the Cadell fault near Echucaon the Murray River (Harris 1939), although
with a throw of a maximum of only 12m, caused
a major diversion of the river. The fault is
across the river (N.-S.) with a tilt downstream
to the W. The main flow of the Murray River
was diverted S. along the fault line, to join the
Goulburn R. above Echuca, thus taking over
some 80 km of the course of that river. Somewater flows N. along the N. arm of the fault
line to Deniliquin, then turns W. as the EdwardRiver to form an anabranch fed through the
offtakes in the low-lying forest areas betweenTocumwal and Echuca. So this minor fault,
because of the flatness of the country, strongly
diverted Australia's largest river, and created
the Echuca Depression in which a lake wasformed, and from the sandy beaches of whichthe Bama Sandhills were built. The N. and S.
divergences of the Murray River join up to
the W. again, thus surrounding the triangular
Cadell Tilt Block. In the mountainous upperreaches of the Murray, the development of afault across the river with a final displacementof 12 m. would have a negligible effect, but onthe flat surface of the sedimentary basin, it
caused a major dislocation (Bowler and Har-ford 1963, Bowler 1967).
The continuing influence of extremely oldlineaments to the present time is illustrated bythe Berridale Wrench Fault (Lambert andWhite 1965), the main movement of which is
dated Early to Middle Devonian, with con-tinuing minor dislocation during the Tertiary,while "seismic activity . . . indicates possiblerecent movement." Hills (1961) has com-mented that "in Australia the results indicate
long-continued activity along ancient lines ofweakness and considerable evidence forgreater mobility of the cratonic units than is
normally attributed to them. Lineaments andblocks are prominent and many of the geo-
NULLA NULLaXsTN.
s.a...a STATION
SH« 1
' 4
H, MURTHO
N.S.W'
N
GLEN ESKSTN
SCALE; i o 2 *
Site' ," % CANEGRASS
' ' SWAMP
JTAREENA
P. CAL tt LAL
JTALGARRY STN.
TAR A DOWN£« STN.
DARLINGi
RIVER
WALKMINGASite 3
yKULCURNAVSTN.CAL L
*b. \wARRAKOSTN Site 5-
^NAMPOO Site 4
pSharp Point
ea Site in
^WARRAN ANGA7 STN
.
DUNEDINPARK STN.
205fl. CONTOUR 62-r
P, ANABRANCH
ROAO>
pS RIVER
FRENCHfvTaNSCREEK
Dftc=>
y WANGLMMAaP. UTAH
S.A.H. PARINGA
LiNDSAT RIVER ^©<?
P, MOOR N
A
Site 13
,Site 12——(l4>
BBfiRIBEE STN. P. WARINAB
Site 19
MURRAY^RlVER
^^MOOF C^>
oNED'S CORNER"
STNrPOTTESWAKAGEE
J
Site IS
^^^tK> L '*X(fg[s3 P205*!. C0NTOLR 02>P. OLN ey
VIC.
LAKE WALLAWALLA<»(> loot ci^KEERA^£TN
^ULNINE SJ
P. WOOLWOOLAN
*fc::: :
: CONTOUR i2- P. TULILLAH
P. MULLROO P. WALLPGjLLA5IURT HlGHW
.,' V
PEDOCAL
-* w / ~"
_3 — 4hJ --<-
—
:—2fi . f \
—*J —:
—
PALEOSOL>. 2",• • \ . / . N
• \- Z. \ - ^ ^— ^
- $ X.
^'
*
/)'''V'
' "**"N
- H* s
\ K.s--<-N\ . \
' -\. /•
v
' \' / \
•
LOXTONSAND
ALEOSOLPALEOSOL
? S /\ / \
/ s. ,' \-S-\'- X •/ • \ • \ / \
/ \ • / \- / \-
N -/* \ \ / \
/ \ / \ / X-\-/\ \ / • s-••%•/ %•/*•%.-• \- / • N \ / \
REDPEDOCAl
PEDOCAL
\ s \.' v / \ / \
\ • \ \ / \
- --..
\ S \
\ / \ \ s* \
/ \ / \ / \
\ .- V \ / \
/ \ / N / \
\ s \ \ s \
' \ / \ / \
\ .'-
\ / \
/ \ / \ / v
v / \
/ \ s \ y \
\ / V \ s \
-' \ / \ S N
\ .' \ \ / S
' *' , '.•jl S. / \
/ \ ,' \
PAIEOSAL PALEOSOL/ N- / . \ / \•%••• \ /• N- /./ < \ - /-. N - / \\ - / \ / N - s .
' \ / - \y\
\ / N/ \ •
LIMESTONEWEST OF5ECTION
i^liUU^
207-27'
PWD BM
PEDOCAL
DOLOMITE
"DOLOMITE
\ / Sf N. /
S. /" N
-- \ ,' v
x / \
/ \- /\ s \
PEDOCAl\ , x
t \ /\ / \
S N /\ /• N
\'/A/• N- / -
• N- / X/ s-/V / V
/ ' \ X •
V - j/ • V,
N.S.WSix MILECREEv
P. NEILPO
(15/ Site 14
IWENTWORTH
mIirray River
to©
Site 17 Mte 16
P YE
MERIDIANROAD
Lta
Site 15
' B OV^-tTR E EK
MAIN ^CHANNELH I G H W A ¥
VICTO " I L D C P *
5A1E0SQL
/ s /\ / \
/ \ /-
s * / ^• \ - /N • / • N
REDPEDOCAL
RED
PEDOCAL
<QN6TQ^
RED
"EDOCAL' \ T~~\
\. S N• v / '
N ' \/ - \ /•% / \
LATERtrE.NOT
SUBVEVEOTO HIVE*
10 12 13 15 16 17 18
Fig. A—Map of project study area, with a series of measured stratigraphic sections. Thenumbers below the sections appear in circles at the sites on the map above. The205 ft. (62 m) contour is the level to which flooding by the Chowilla Dam wasdesigned. Such contours are on the S.A. datum of M.S.L. Adelaide = 106-88 ftLake Wallawalla is dry, and Lake Cullulleraine is an irrigation storage. Only sections1-2, 5-6 and 17 reach the local plateau level. Behind the other sections, the terrainrises, and m many instances Blanchetown Clay was proved there.
Note that the downwarp in this area from the Pinnaroo Block to the W. causesthe Loxton Sand to disappear below river level as one moves E., and then similarlythe Parilla Sand. The latter reappears at Paringi far to the E. In the Lake Victoriasyncline, the Moorna Formation is intercalated below the Blanchetown Clay/ChowillaSand complex. The vertically hatched areas on the river floodplain are Rufus Forma-tion. The base of each section is river level, while No. 6 on Lake Victoria is at
the level of Lock 9, from which the lake is filled. H (in S.A.) = Hundred; P (in
Victoria) — Parish. Section 18 (not on map) is from Paringi, SE. of Mildura.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER
logical basins are framed by lineaments". Asalready indicated, this is true of the MurrayBasin.
4. Structure Pattern at Surface
Firman (1970) has recorded a faint pattern
of structural elements which he recognized fromlow-flying light aircraft and mapped fromaerial mosaics. The results were related to
geophysical and geologic information. Theywere used to define the Pinnaroo Block (Fir-
man 1966).
5. Degree of Tectonic Control of Sedimenta-
tion
Without subscribing to a complete theory
of tectonic control of sedimentation (e.g. as
discussed by Crook 1967), it can be stated as
fact that the tectonic style of the area affects
the nature of the sedimentation. On first study-
ing the area, I was impressed with the almost
universal occurrence of fine sands—in the
Parilla Sand, the sandy phases of the Blanche-
town Clay, the Chowilla sand (although a
channel deposit), the various Quaternary
riverine and dune systems, and much of the
Moorna Formation. It was with some surprise
therefore that I came upon the ossiferous
gravelly sand of a channel deposit in the
Moorna Formation at Fishermans Cliff, onMoorna Station. Even the sediments of the
present Murray River, rejuvenated by incision
of a course round the Pinnaroo Block, are
fine clayey sands on the floodplains as seen in
the red outliers (relicts of a Pleistocene terrace
with giant marsupial remains, here named the
Rufus Formation ) and the Coonambidgal
Formation. Washed sands characterize the
channels. This general fineness of sediment is
a function of the low dynamics of the streams
which in turn is due to their exceptionally low
grades, a result of the mild tectonics. See grain
size analyses in section on sedimentation.
The tectonic uplift of the Pinnaroo Block
dammed the Murray River and brought about
the formation of the ancient Lake Bungunnia
(Firman 1965). Tectonic movement was there-
fore the initiator of Blanchetown Clay deposi-
tion.
Flatland Geomorphology
Sturt (1849) began his account of explora-
tion in Australia with this statement: "TheAustralian continent is not distinguished . . ,
by any prominent geographical feature . . .
nor do any of its rivers . . ., not even the
Murray, bear any proportion to the size of the
continent itself. This description is par-
ticularly true with reference to the area nowdescribed.
Rivers
Meanders apart, the River Murray in the
study area has a general direction of a little
N. of W. To the W. of the Darling Anabranchthis general direction is offset about 20 kmby a section of the river that flows SW. Thereason is not known, but it is suspected to be
associated with tectonic influences that also
caused the Darling Anabranch to gain entry
to the Murray where it does. The locations of
Wallpolla Creek with its anabranches and BoyCreek may be connected with this same pattern.
The designers of the Chowilla Dam envisaged
(at one stage at least) filling to the 205 ft.
contour (62-5 m). This contour is based on the
River Murray Survey of 1909-1914 when the
South Australian datum was used, which is
approximately 105 feet below Mean Sea Level.
This unusual datum originates in the fact that
part of South Australia (the Lake Eyre region)
extends below sea-level. Thus the 205 ft. con-
tour is approximately only 100 ft. (30*5 m)above M.S.L. although so far from the sea
(Wentworth is about 820 km by river fromthe sea). This contour is relatively close to the
river at the S.A. border but further E. swings
out widely both N. and S. of the river. It nears
the river again at the S. end of this SW. flowing
sector, and follows it fairly closely to wherethis sector begins. The position is complex in
die Wallpolla Creek area because this stream
is actually an anastomosing system of used andabandoned channels, as though tectonically in-
stigated shifts of channel had repeatedly oc-
curred. Perhaps the curiously isolated, but
large channel called Boy Creek is a former
course of the Murray River. As our project
had to cover so large an area in so brief total
10 EDMUND D. GILL
field work time, our investigation had to be of
a reconnaissance nature, but in numerous places
in this account interesting problems for future
investigation (such as this one) will be in-
dicated. Some are listed in Appendix 2.
The Darling River enters the Murray bytwin streams—the Darling itself at Wentworth(PL 1 ) and the Darling Anabranch about 14 5
km further W. Anabranch is an Australian
word and appears to have been used first byColonel Jackson (1834, p. 79), who states
"Thus such branches of a river as after separa-
tion re-unite, I would term anastomosing
branches; or, if a word might be coined ana-
branches/' Thereafter, it was used by Leich-
hardt (1847,2: 35), Start (1849, 1: 93),Blan-dowski (1857, p. 174) and others. The wordwas spelt without the hyphen by Leichhardt,
and I presume was originally used by Jackson
only to indicate that he had taken the ana fromanastomosing and joined it to branch. Some-times the first syllable is spelt anna (first byBlandowski 1857, p. 128), which is incorrect
by derivation, and sometimes as two words (as
on the military map) which is not according to
original usage.
The Darling and its Anabranch arc both widestreams. Twin streams are like sympatric species
of animals—one usually has a slight advantage
over the other and in time takes over. How-ever, both these streams flow strongly at pre-
sent and a comparative study would be an in-
teresting project. A string of lakes, filled only
in flood time, is associated with the Darling
in the Menindce district, while another groupthat could be regarded as belonging to the
same system, is associated with the Anabranch.The origin of these lakes is unknown, but mayhave a similar genesis to that of Lake Victoria
discussed later in this paper. Lake Menindeehas yielded a rich deposit of fossil vertebrates
(Tedford 1967), and more recently LakeTandou further to the SW. has yielded a similar
fauna, an account of which is provided in this
Memoir by Dr. D. Merrilees. Further E., a
similar system of dry lakes includes LakeMungo, which has yielded records of bothmarsupial and human occupation plus evidence
of high lake levels (Bowler et al. 1970).
The so-called creeks of the area are generally
dry, and their significance is in the past rather
than the present. The stratigraphy of the river
flats at the Chowilla Dam site show a marked
change in river regime. Six Mile Creek between
the Darling and its Anabranch is an old river
channel, as also probably is Boy Creek. All the
creeks, and indeed the rivers themselves, are
classifiable as misfit streams (Dury 1960). Theamount of water that flows is small compared
with distance traversed and the size of the
valleys. It is important not to think in terms
of the present flow, for the streams (especially
the Murray) are controlled by dams and locks.
Under natural conditions floods and droughts
had severe effects. As recently as 1945 I walked
dry shod across the bed of the Murray River
near Koondrook. George Everard in an un-
published diary tells of a low river level in 1857
observed by him between Kulkyne and Went-worth. "We could cross quite easily in mostplaces, having to swim the middle about a
chain; the rest we could walk, the water being
so shallow1'. Thus, under natural conditions the
the river at times was not much of a barrier,
whereas at present a high water level is always
maintained.
The junction of the Murray and Darling
Rivers (where the town of Wentworth stands)
is an historic site, the area having been visited
in:
1836 by Mitchell (1839, 2: 116)1844 by Sturt (1849, 1: 103)1853 by Mueller (Barnard 1904)1856 by Blandowski (1857).
Rainfall and Floods
The rainfall is only 22-26 cm in this area,
contributing little to river flow because it is so
low, and because surficial sands soak up muchof it. Some rain runs into enclosed areas whichmay result in salt lakes such as occur E. of the
Lake Victoria dunes on Talgarry Station, andSalt Lake E. of the Wentworth-Renmark roadon Dunedin Park Station. Heavy local falls mayresult from thunder storms. During our field
work at Lake Victoria, one fall of 9 cm oc-
curred, about 40 per cent of the average an-
nual rainfall. The evaporation is about 18 m.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 11
Flood water has two sources—Summer mon-soonal water that flows down the Darling (Sturt
1849, 1:6), and Spring meltwater from the
Kosciusko highlands further S. Normally, the
floods do not coincide (as noted by Mitchell
1839, 2 : 116) but from time to time this
happens, and king floods result. In the townof Wentworth a monument stands to the menand machines that saved the town during such
a flood. The oscillation of flood and drought is
a characteristic of this river system which has
been present for a very long time, because the
whole biotic complex is adapted to it. This
characteristic ruined the boat trade, which wasunable to depend on access. An extreme in-
stance is that of the boat that went up the
Darling to Bourke, but could not return for
three years because a prolonged drought so
lowered the water level that passage was im-
possible (Cameron 1966). In 1968 there was
a flow in the Murray of 15,000 cusecs (35,000
1. per sec.) but in 1967 only 780 (1,800 1.
per sec.) during a drought. However, in the
major flood of 1956 the flow was 140,000
cusecs (330,000 1. per sec). The largest re-
corded flood is 156,000 cusecs (368,000 1.
per sec). There was a Great Flood and rain
in the Darling in 1863-4 (Hardy 1969, p. 134)
that brought many hardships, but also green
feed to an arid land. Such events could not be
depended upon by setders. For example, Tate
(1885) records that floods filled Lake Mombain 1864 then not again until 1885. Cameron
(1966) describes the big flood of the Darling
River in 1890, and the unsuccessful attempts
to save the town of Bourke.
Drought (Heathcote 1969) in Australia "is
a permanent feature and not just a misadven-
ture". It has to be planned for as a cyclic event.
For some time this was not understood. For
example in 1883 the Government was doubtful
(Everard, n.d.) whether the N. part of Victoria,
called The Mallee (Hardy 1914) "was worth
saving". When Sturt reached the Darling River
on Feb. 4th, 1829, at New Years Creek, he
found the water too saline to be potable, and
the Aboriginals had left the area.
River Terraces
Between Wentworth and the Chowilla Dam
Site, the floor of the Murray River tract is
exceptionally wide (Fig. 4). This is probably
due to the river oscillating from side to side
during the period of impedence while the
stream was incising the Murray Gorge in the
Pinnaroo Block. A local tectonic factor is also
probably involved, as will be discussed in the
section on the origin of Lake Victoria. It is the
broadness of the river tract here that will causeso large an area to be inundated (1300 km2
}if a dam is built at Chowilla, and it is the usually
steep walls of the valley tract that nevertheless
will contain the water.
The river tract is an uneven pattern of dull
gray and bright red. The gray alluvium is
mostly, but not all, covered by the king floods,
so some of it was deposited by still higher
waters. However, the surface is relict (except
for the lowest areas ) and consists of old
channels of well-washed sand and floodplains
of clayey sand, which have not been maskedby any later alluviation. These are comparablewith the Coonambidgal Formation described
by Butler (1958), Pels (1966), Bowler(1967), Firman (1967) and others. The gray
country consists of uncompacted sediments
that may be treacherous for the motorist in
wet weather. The red country is compacted andpassable in wet weather. The gray country is.
unoxidized, while the red is strongly oxidized.
The gray country has little development of soil
profiles, while the red has a well-developed soil
profile with mobilization of carbonates in
earthy patches to small nodules. The gray
country is young, while the red country is mucholder.
Few fossils have been found in the gray
country, but radiocarbon dates therefrom are
Holocene. The red country has provided the
mineralized bones of extinct marsupials, whichelsewhere in this region are Pleistocene. Thegray country is extremely flat and related tothe baselevel of present deposition. Blandowski
(1857, p. 127) was impressed with the flat-
ness of this country describing "clay flats ofremarkable evenness". In Brown's Paddock onKeera Station on the S. side of the MurrayRiver, a survey from auger holes to a benchmark crossed about 3 km of gray floodplain
12 EDMUND D. GILL
and found only about 7 cm difference in level.
On the other hand, the red country consists
of the remnants of a Pleistocene floodplain that
stands well above (commonly 7 m) the exist-
ing floodplain (Fig. 5 ) , and has suffered
erosion, sometimes extensive. In the section
on stratigraphy, this formation is described as
the Rufus Formation, and some account of its
sediments given. The Coonambidgal Formationis characterized by Eucalyptus largiflorens with
E. camaldulensis along the waterways. TheRufus Formation is marked by non-myrtaceous
plants such as Oldman Saltbush, Atriplex num-mularia, Callitris and Casuarina. The great
variety of the shapes of these outliers catches
the eye on the air photos, but those checkedproved to owe their morphology to the direc-
tions of the Coonambidgal channels. The small
area covered by Rufus Formation outliers in
Fig. 5 is a measure of the extensive erosion ef-
fected by the Coonambidgal streams. The
strong contrast between the Coonambidgal
Formation and the Rufus Formation in oxida-
tion, compaction, pedology, erosion, secondary
mineralisation and in palaeontology, indicates a
considerable time break with consequent change
of environment. It would be helpful to knowhow great this period is. No dependable chron-
ology is yet available. There are indications of
the order of age in that the lunette on the E.
side of Lake Victoria onlaps the Rufus For-
mation there, which contains Sarcophilus, nowextinct on the mainland. The oldest date fromthe Lake Victoria dune system so far is 17,530yr (ANU-404A) but this is high in the suc-
cession so the base of the dune is considerably
older. At Lake Mungo (Bowler et al. 1970),in this same river system, the base of the dunesystem is of the order of 32,000 years.
The deep weathering of the Rufus Formationwould take a great deal of time. Norris (1969)says the process of reddening "is promoted by
PLATEAU
Chieity Bkirn-helouii Clay with surtici.il sand oriented casl-wesl
Fig. 5—Trace of air photos to show the distribution in the river valley (N. of theMurray River and E. of Lake Victoria) of the higher red country (RufusFormation, oxidized remains of a Pleistocene fluviatile terrace) and the lowergreenish gray country (Coonabidgal Formation). Sites 9-10.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 13
warm temperatures, oxidizing conditions andthe periodic presence of moisture. Reddening is
due to the gradual weathering of iron oxide andsilicate mineral grains; the weathering tends to
cause the coatings to become thicker and af-
fects increasing numbers with the passage of
time. In addition, evidence suggests that the
haematitic grain coatings are resistant to the
abrasion associated with aeolian transport".
Palaeoclimatic evidence indicates changes in
river regime with time in the Murray valley,
and it will probably be possible later to corre-
late this stratigraphic break between the
Coonambidgal Formation and the Rufus For-
mation with one of those periods of change.
The Rufus Formation may be middle Pleisto-
cene in age.
Meanders and Oxbow Lakes
"All the varied forms of the land are de-
pendent upon , . . structure, process and time".
W. M. Davis 1954. In the area studied, the
Murray River is a mass of meanders, and ox-
bow lakes (called billabongs in Australia) are
very common (PL 3). This exceptionally ex-
tensive meander belt is worthy of closer study.
In the short time available, the oxbow lake be-
tween Moorna Homestead and Woolshed close
to the Wentworth-Renmark road (upstream
from Fishermans Cliff and the paddock called
The Selection) was studied briefly. This at-
tractive body of water has a high biomass pro-
duction. Some 60 species of birds were counted
on or around it. On one occasion about 200
black swans alone were counted. Fish, tortoises,
atypid prawns (determined by Dr. W. D.
Williams) molluscs and ostracods were noted
without a survey being made. Such places
formed an important source of food for the
Aborigines. This high biologic production re-
sults in carbonaceous sediments on the lake
floor. When a drought was on, the lake level
dropped and a spade hole was dug in the peaty
floor at the S. end of the W. arm of the oxbow.
It revealed only 23 cm of carbonaceous sedi-
ment over a highly compacted mottled clayey
sand which is apparently the B horizon of a
former soil.
The cliffs of the oxbow are very steep and
the talus slope reaches only 2-0 5 of their
height. Where gulches debouch into the billa-
gong, there are small deltas covered with reeds.
Thus the thickness of floor sediments and the
morphology of the cliffs suggested a youthful
age.
Figure 6 provides a surveyed section of the
N. side of a gulch cut in the NW. cliff of the
oxbow lake. This traverses the delta of sediment
washed out of the gulch. The rainfall is only
23 cm p.a. but much of this often falls sud-
denly in thunderstorms and so erodes with
high energy. Examination of other cliffs in the
area showed this to be a widespread condition
and hinted that there might have been a "1000
y. flood" or such event (or series of events)
that cleaned out the river courses and oxbowsby flooding of unusual energy. It was therefore
decided to bore the floor of the oxbow lake to
obtain sediment from the base of the "clay
plug" for radiocarbon dating. If successful,
this would give the order of age of the present
geomorphic unit.
Mr. W. Levy kindly provided a boat. TheMuseum carpenters constructed a high tripod
to stand on the floor of the lake to which a
suitable boring apparatus could be attached.
However, some 25 cm of very soft gray sedi-
ment made it difficult to set the tripod, the
legs of which had to be heavily weighted to
make them stable. C.S.I.R.O. Division of Ap-plied Geomechanics, through its Chief, Dr.
G. D. Aitchison, kindly supplied the boring ap-
paratus and the services of Mr. Gavin Renfrey,
without whose expertise we could not have
mastered the many problems that arose. Ul-
timately, the core was taken, the profile being
Rid Paloual
Carbonate fraebuks
BL^M i[FT(»\\\ t LAY
Delta of ouiwusli from gulch
Water level in mbow lake jt lime ot survev. Height ot station 12>2
Fig. 6—Surveyed section, Moorna East oxbow, W. ofWentworth, N.S.W. Site 13.
14 EDMUND D. GILL
Gulch (Fig. 6)
\s^ ^?r SluW N - llul
v BORE SECTION
•
BoreI : ra Water
— —OXBOW I \K1- - 7 (] (.'III
—j- Dark
-~
-gryj d;i\ _— —
1 I'oa-sk-J 1 loodpbm 1
» 12-7 .ni
-Ciru> Any -
CM Sample*• — —
|
°o
(. canpuct
Mottled
l'alcu,nl
y^^ MURRAY R1YUK
Fig. 7—Moorna East oxbow with the gulch besidewhich the section (Fig. 6) was surveyed;also site of bore. On right is section of bore.Radiocarbon date 770 ± 150 yr B.P. Site 13.
as in Fig. 7. The sample from the layer im-mediately above the eroded former land sur-
face was dated by Professor K. Kigoshi as
770 ± 150 y. B. P. (GaK-3217). The carboncontent was not as high as anticipated (thepercentage fell off at the base) so that thesample was insufficient for a precise dating.
However, the order of age was determined, andthe juvenility of the oxbow lake established.
This was all that could be done under the cir-
cumstances, but the further elucidation of this
matter would be a worthwhile project.
Plain
The plain in which the river tract is incisedis of the order of 36 m above the river (PL 2).The plain descends to the river or its flood-plain usually by steep slopes or cliffs. From theair the green linear river tract contrasts strong-ly with the general redness of the terrain. Tothis natural ecology is now added the greenpatchwork of the irrigation blocks. A questionworthy of investigation is why the MurrayRiver has followed the course it has across
this flat plain of lacustrine (Blanchetown Clay)
and riverine sediments. As already pointed out,
there are tectonic factors involved. There are
also lithologic factors. It is noteworthy that a
marked difference exists between the country
S. of the river and that to the N. The Victorian
section of the plain is characterized by the Big
Desert and associated semi-arid sandy country,
all geomorphically distinct by reason of its E-W. longitudinal dunes with sharply defined
crests (air photos). These dunes have earthy
carbonate subsoils, and calcrete nodules are
only found (in my experience) under these
dunes as quarries and borings showed. By con-trast, the country N. of the river has relatively
only a veneer of sand, and over large areas
Blanchetown Clay outcrops at the surface. Thus,conservation of water S. of the river is a prob-lem, while N. of the river large tanks excavatedin the green Blanchetown Clay are common.As the air photos readily reveal, the country N.of the river has an E-W. grain with some de-velopment of longitudinal dunes, but the usualcarbonate soils are those bearing laminatedcalcite nodules in the subsoil. As these twoareas belong to different States, this importantdifference has not been indicated in studies
published so far, and merits further investiga-tion.
Because the Blanchetown Clay is of UpperPliocene or Lower Pleistocene age, its surfaceis a relict landform. So also are the land sur-faces consisting of Chowilla Sand, and furtherE. of prior streams (dating >40,000 by radio-carbon). Younger relict surfaces are dunes in
the Mallee with uppermost subsoils dating16,000-18,000 years (the substrate is of courseolder than the soil on it), and the sedimentsof the ancestral streams further E. that are notburied by later deposits.
Thus the broad flat plain of the MurrayBasin is largely a relict surface of varying agesbecause
(a) The river system is confined to a tract cutinto this plain, and
(b) The dune systems are mostly dead, orjust being re-modelled.
The relict nature of the surface is also in-
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 15
dicated by the polygenetic soils. The ground is
packed with carbonates, which will be discussed
(including their chronology) in a later section.
Relict surfaces imply terrain stability, but at
the present time the countryside is characterized
by instability. Gulches in active erosion charac-
terize river banks and the sides of Lake Vic-
toria. Much of the surface of the dune system
on the E. side of Lake Victoria is now mobilewith extensive blowouts and sandfalls. Thefence between Nulla and Talgarry Stations onthe E. side of Lake Victoria overlies three
others because of the build-up of sand. Blow-outs among the saltbush are common and a
great number of such bushes have piles of
loose sand under them. During the drought,
dust storms were common. The reconciliation
of relict terrains with all this instability is that
the latter has mostly occurred since Europeanoccupation (as will be discussed later).
Finally, this review of planar areas abovethe floodplains should refer to some bordering
flatlands or remnants of them along the river
channels. At Cal Lai (Fig. 4) one descends
from the general plain to a flat area well above
the floodplain (e.g. behind the police station).
This flat area (obviously once more extensive)
is due to differential erosion, as the area cor-
responds with the top of a silicified zone that
is part of a paleosol. The same kind of surface
at the same stratigraphic level occurs on the
E. bank of Salt Creek, Kulcurna Station, for
some distance.
On the W. side of Lake Victoria there are
mesas and promontories that suggest a former
surface below the level of the general plain. It
may represent a stage in the excavation of the
Lake Victoria basin.
Salt Lakes and Swamps
When rain falls on the flat plain of this
country, much of it lies about in claypans or
penetrates dunes and sandspreads where it
constitutes a temporary aquifer, apparently
made use of by the Aboriginals. In other places
the water runs into hollows with no outlets and
salt lakes develop. Such depressions are not
uncommon, but their origin has not been in-
vestigated.
Salt lakes have been noted in two types of
geomorphic situation, viz.:
(a) In depressions, e.g. on Dunedin ParkStation E. of the road from Wentworthto Nulla Station; also on Pine CampStation.
(b) In negative areas where drainage has beencut off, such as behind the dune system at
Lake Victoria. A salt lake occurs E. of the
dunes on Talgarry Station S. of the track
to the pump at Lake Victoria. To bore this
lake, follow changes in regime, and byradiocarbon date its time of origin andhistory was a project that had to be aban-
doned for lack of time. As the dune system
is apparently within the range of radio-
carbon, presumably the lake is also, since
it was formed by the dune system blocking
the drainage.
Fletcher Lake NE. of Wentworth (Milduramilitary map, 2 cm=5 km) has a high lunette
on the E. side. There are many dry lakes suchas Lake Wallawalla (Fig. 4). Lake Culluller-
aine is a water storage.
Cane grass swamps are common in the flood-
plain of the river system, a typical example be-
ing one on Moorna Station we called Triple
Swamp by reason of its trilobed shape as seen
in the air photos (Wentworth Run 2, 1363-
5231, 9-7-65, NE. of C.P.) Stratigraphic sec-
tions from the cliff behind this swamp are des-
cribed later.
In the NE. part of Lake Victoria basin be-
hind the dune system E. of the lake is a canegrass swamp. An auger hole was sunk about
20 m E. of the track that passes it, and traversed
on 24th July, 1969:
00 -0-3 m Reddish brown (Munsell 2-5 YR 5/4)clayey sand gradually merging to
0-3 -0-63m Light brown (7-5 YR 6/4) sand withhard carbonate
0-63-2- 13m Light yellowish brown (10 YR 6/4)slightly clayey sand without carbonate.Moisture at 2m and so a little darker
2- 13-2- 59m Mottles of light gray (5 YR 7/1). Be-comes greenish gray (5BG 7/ 1 ) at
2-31m. Pyrolusite at 2-53m. Heavy at
2-61m.2 -59-3 07m Very pale brown (10 YR 7/3) sand.
Darker material included 2 -74-3 -04m.3 07m -f- Light gray mottles again. Salty water
at 3 -35m.
16 EDMUND D. GILL
Claypans (small playas)
This term is widely used in Australia but,
although related in meaning to "pi aya" (Chico
1968), is not its equivalent. The term playa
is used in many senses, but in Australia is ap-
plied to a large dry lake (Jutson 1934), while
"salina" or "salt lake" (e.g. Bettenay 1962)is employed if a salt crust is present. A claypan
is a small flat unvegetatcd area, in an arid
or semi-arid environment characterized bysheetflood deposition of fine sediments (usually
red to pale brown). In the area studied clay-
pans are commonly 20 to 30 m in diameter andtend to be round in shape. On Nulla Station
some large claypans were joined up to forman air strip for light planes in fine weather.
By extension, the term claypan is used also
in the study area for places of similar appear-
ance to the above, which arc in reality places
stripped by wind erosion of the surficial sand
to expose a flat hardpan of similar sand boundby clay skins. Such sediments do not effervesce
on application of acid, and so the amount of
carbonate is negligible. On such stripped zones
Aboriginal implements and rock debris form a
lag deposit; also lumps of baked clay that wereused as cooking stones (to be discussed later).
Claypans can form in dunelands by centripetal
wash of the fine fraction which has cither beenblown up into the dunes as aggregates or de-
posited on them by aerial transport. Fossil clay-
pans of this type occur in the dune system onthe E. side of Lake Victoria. They arc charac-
terized by fine herringbone rills when exposed.
Two important features of the geomorpho-logy remain for comment, but each is a majortopic and so individual sections will be devoted
to them, viz.:
(a) Dunes(b) Lake Victoria.
Dunes
The semi-arid Murray Basin, with its ex-
tensive fluviatile sand sources and a wind re-
gime with seasonal persistent westerlies (Fig.
8), is a natural setting for duneiields.
The Murray River at Echuca adopts a NW.trend which it maintains overall as far as S.
SURFACE WIND ROSE FOR MILQUR*
P.r.od: 1967 1971 inclwai**
Ofcs Houn: 3 om, 6 om, 9 om,
noon, 3 pm, 6 pm, 9 pm
WEST 3
Peitentnge frequency
12 3 4
Wind speed [kno(s)
6 7- 16 17 2/
Numbei ol obs ol winds ?8 33 knots
in the 5 ye.if permd
©wim.1-. ,nf from directions shown,
dfiiirir within circle is percentage
nmntm uf caltns
Fig. 8—Wind rose showing the predominant W. windsthat determine the direction of the longi-tudinal dunes. Prepared by The Common-wealth Bureau of Meteorology.
Australia. This is also the orientation of thebroad ridges that constitute the chief geo-morphic feature of the Mallee in N. Victoria(Hills 1939, Thomas 1952, Lawrence 1966).Lakes and other surface features (and as aresult the railways) are affected by this generaltrend, but it has no relation to the configurationof the bedrock below the Murray Basin sedi-
ments. Across this lineation are the E-W.longitudinal dunes of the surficial WoorinenFormation (Lawrence 1966, Fig. 8). In someplaces the dunes spill into lakes, proving their
longitudinal character and direction of sandmovement.
Following the Murray river to the NW. fromEchuca is a large very flat floodplain (of both
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 17
present and past) consisting of very clayey sedi-
ments and without ridges. NW. of Benjeroopthe flat plain ceases, and areas of red soil with
hard carbonate nodules rise like islands abovethe gray sediments (e.g. near "Winlaton" home-stead). It may be that the Leaghur Fault of
Macumber (1969) at its north end swings
round the N. margin of the Gredgwin Ridge to
this area, and is partly responsible for this
change in terrain.
The broad NW.-SE. Mallee ridges represent
the interfluves of rivers once active, while the
E-W. longitudinal dunes represent the drying
up of the rivers and establishment of a desert.
Thus Mabbutt (1968) remarks about Central
Australia deserts—"As in most of the world's
deserts, aeolian sand surfaces mark the retrac-
tion and disintegration of drainage systems and
the abandonment to wind action of coarse-
textured alluvia in former depositional en-
vironments".
The evidence of former rivers signifies morehumid conditions (at least in the areas of
origin), while the establishment of the dunes
proves drier conditions. That the dune system
is now inactive except for some of the moresandy crests, that they are almost completely
vegetated, and that they commonly have a
fairly deep soil developed on their surfaces,
proves that conditions are less desertic than
when the dunes were formed, i.e. the climate
has been both more humid (rivers) and drier
(desert) respectively.
The dunes are parallel, which is "character-
istic of large open expanses of dunes where
sand movement has been least complicated byrelief or drainage" (Mabbutt 1968). Theclimax of dune building was when this system
was active. Mabbutt and Sullivan (1968) dis-
cuss whether the dunes of the Simpson Desert
are normal dunes or windrifts (swales formed
by corrasion). They support the former view
but perhaps these ideas should not be set in
apposition, as at least some of the dunes in the
Simpson Desert have cores of older material,
as do some in the Mallee. As subsequent
evidence will indicate, the Mallee ridges are
polycyclic as observed by Hills (1939). Jen-
nings (1968) notes that desert dunes cover 19
per cent of Australia. His map shows an Aus-tralia-wide response in dune trend to the
continent's overall wind system. The directions
of the dunes in our study fit the continental
pattern of wind directions. It is likely that during
the Last Glacial at certain times the Mallee
with its active dunes was comparable to the
present Simpson Desert. For more information
on the Simpson Desert see Folk (1969, 1971)and Boyland (1971). Thus there has been a
range of regimes from river country to dust
bowl.
Like the savanna lands of other parts of the
world, those in Australia reflect the climatic
variations of past ages. They are marginal
areas that more readily register climatic shifts.
Thus Thomas (1971) says, "Thick duricrusts
beside the River Niger near Niamey wereformed under humid conditions in the Tertiary
period: sand dunes are about 20,000 years
old but are fixed". "Chad Basin in W. Africa
provides evidence of former climates. Duringarid phase over 20,000 years ago, seif dunestrending NE. to SW. were formed. By 10,000years ago wetter conditions had fed enlarged
'Mega-Chad' .... In the last 7,000 years
the climate has come drier and Lake Chad has
shrunk. Transverse dunes appeared during
minor climatic changes".
Mallee Dunefield
Hills et al. (1966) state that "Longitudinal
dunes are best developed where there is anabundant supply of sand, especially in large
alluvial basins". The Mallee Dunefield of E-W. longitudinal dunes is a re-organisation of
sand from the Diapur Sand, Chowilla Sandand such, during dry periods. The dunes exist
because there was an adequate supply of sand to
windward, and the aeolian traction was suf-
ficient to move the sand, but not strong enoughto cause only deflation (as on a gibber plain),
leaving a lag deposit. The climate of the time
was dry enough for the sand to be mobile, andfor the vegetation to be depressed enough to
prevent immobilization of the dunes. The dune-
field is extensive even more so than is shownon the World Aeronautical Chart, or on the
map in the Reader's Digest Complete Atlas of
18 EDMUND D. GILL
Australia (1968). On the aerial photos, the
dunes are very distinct and the crests appear
sharp, but on the ground they appear rounded.
This is of course a matter of scale. As the field
has long been immobile, the crests cannot
remain sharp as in an active dune system. Thefirst two years of our observations were a time
of drought, and this field was also observed
by the writer during the severe 1945 drought.
During the latter, the country was denuded of
green herbage, and even the domestic geese
flew round of an evening (usually they are too
heavy). However, the dunefield was not re-
activated. The crests in more sandy areas be-
came mobile but the system as such remained
inactive. The dunefield is thus definitely relict.
However, we should think in terms of the
natural conditions that existed in the Mallee
before European occupation, because the latter
has greatly increased sand movement. Thewestern Mallee is less stable than the eastern
part. Some idea of the Mallee in its original
state can be obtained from early observers as
recorded, for example, by Blandowski (1857),
Bride (1898), Everard (n.d.), Hardy (1969),Hawdon (1852), Kenyon (1942), and Morton
(1861).
1. Swan Hill Paleosols
Churchward (1961, 1963 a-c) studied these
dunes and their soils in the Swan Hill area onthe Murray River. As a pedologist he was con-
cerned chiefly with the soil mantles, and his
findings can be summarised thus:
It is to be noted that with increasing age the
soils are more mature, therefore one would
expect each pedogcnic phase to be successively
longer. The increasing maturity of the soils
is shown by the depth of carbonate leaching,
the degree of clay segregation, the pedality, the
amount of development of lustre on the ped
faces and compaction status. Each of these
factors involves time. However, the length of
time between the dates for the Speewa and
Kyalite soils is not less than that between the
dates for the Byamue and Speewa soils, but
47 per cent more. No simple damped harmonic
curve can therefore express the history onpresent evidence. The reason is that two un-
related factors are involved, viz. the accumula-
tion of the substrate (dune building), and the
modification of the substrate surface by pedo-
genic processes (soil formation). The former
is an unstable phase of the terrain, and the
latter a stable phase. Oscillation of these states
is indicated by the successive dune deposits and
soils. However, the diminishing intensity of
soil formation raises the question as to whether
widespread palaeoclimatic changes are involved
or local variations. This can be checked by ex-
amining what has happened in other parts of
of the dunefield. During the present Project,
such checks were made. Multiple paleosols werefound to characterize the dunefield. Twopaleosols were the commonest, but three andsometimes four were also found. Sections at
three widely spaced localities are given in the
SOIL MANTLE LIME CLAY PEDALITY PED FACES COMPACTION4. KYALITE(15,550 yr)
3. SPEEWA(24,000 yr)
No mobilisation
At 35cm
No segregation
MovementAtoB
None
Weak
None
Low lustre
clay skins
Soft granularstructure withorganic buildup
Firm
2. BYAMUE(29,750 yr)
At 120cm Strongsegregation
High Thick glossyclay skins
Dense
1. TOOLYBUCKNom—(Oldest) becoming
acid
Very strongsegregation
Veryhigh
Very glossyclay skins
Very dense
The dates provided have been published bythe Australian National University in Radio-
carbon vol. 12, and the laboratory numbersfrom youngest to oldest are ANU-183, 184,
185 respectively.
following paragraphs with radiocarbon datings
of the soil carbonates. The reliability of suchdates will be discussed in the section onpalaeopedology, and the climatic phases in thesection on palaeoclimatology.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 19
2. Ouyen Paleosols
In 1939 Hills drew attention to evidence of
this kind in the same dunefield N. of Ouyen at
and near the 290 mile post. Carbonate layers
on the Calder Highway (p. 315, PI. 24, fig. 2)he interpreted correctly as paleosols and in-
dicators of more humid periods during the time
of dune building. On the same highway 1 kmS. of the 283 milepost (about 3-2 km N. of
Ouyen and 4-8 km S. of Kiamal), there is a
site on the W. side of the highway about 90 mS. of a branch road on the E. side where I
studied the roadcut shown in Figure 9. Thesection comprised:
06 m Uncompuulcd well*sorttfU(wvll-nwmW! ycLIowisl) iv..i
c. 5YR 5/6 saml
.A Hori/on. ..
ILLLUJJtBH
Road C uttins
5-2 cm Firm sum! 5Yft 5/H
17.8 cm Hard white earthy carbonate
UPPER PALLOSOL • \ Rill erosion
' 0-76 m Very firm sand with atrbonale aiul cluy eemeni
•Nu A Horizon17 8 cm Hard white earthy
UfmriirmiiiuiiuioriAm-LOWER PALEOSOL' .
\". .
'
•'
• 15*2 Very firm clayey sand mottled yellowish-red 5\\< 5, 6 and red
imaumum 2-5YR 4/6)
W Road
Fig. 9—Roadcut about 32 km N. of Ouyen, N.Victoria, on the Calder Highway, showingcarbonate paleosols in E-W. dune (WoorinenFormation). Carbonate from the upper pale-
osol dated c. 16,400 yr B.P.
60cm Uncompacted well-sorted well-rounded aeo-
lian sand of variable colour but aroundyellowish-red 5 YR 5/6.
15cm Yellowish-red 5 YR 5/8 firm sand
c.lScm Hard white carbonate of earthy type (not
solid crystalline) of variable thickness along
the outcrop, grading into
69cm Very firm clayey sand with some carbonate
and clay cement, characterized by rill erosion,
c- 15cm White carbonate similar to the horizon
above90cm Very firm clayey sand, mottled yellowish-red
(5 YR 5/6) and red (to maximal 2 5 YR4/6) with rill erosion.
The two carbonate layers are B horizons.
The clay in the dune sands is interpreted as
windblown dust washed in by rainwater (cf.
Gill 1966, p. 556). Even at the present time
much fine red dust is blown over the Dividing
Range to S. Victoria (Chapman and Grayson
1903) and E. to the Snowy Mts. and beyond
(Walker and Costin 1971).
From the upper carbonate layer a sample
was collected, which yielded a radiocarbon
date of 16,400 +**J
yr B. P. (GaK-3218).
This is comparable in age with the date of
15,550 yr at Swan Hill for the Kyalite soil.
As soils take a good deal of time to develop,
and are not perfectly synchronous over their
range, such variations in age are to be expected.
See also the discussion of carbonate dates in
the section on palaeopedology.
3. Hattah Paleosols
On the E. side of the Calder Highway (79)
between mileposts 309 and 310 (from Mel-
bourne) there are extensive pits for the re-
moval of carbonate soil nodules for road works.
On the S. side of the entrance track, a roadcut
sections an E-W. dune. As pits with nodules oc-
cur to the E. of the cutting on the axis of the
dune, and no nodules are found in the dunes,
it was surmised that the nodule bed underlies
the dunes. To test this an auger hole was put
down on the N. slope of the dune near the
road and overlooking the access track. Thesection proved was:
20cm Reddish yellow 5 YR 6/6 fine sand merging to
18cm Light red 2-5 YR 6/6 fine sand merging to
63cm Light red 2-5 YR 6/6 fine sand merging to
43cm Fine sand with earthy carbonate as small firm
lumps up to 6cm diameter. Colour lighter
because of carbonate, on decrease of which6/6 then 6/4 (light reddish brown) grading to
5 1 cm Reddish-yellow 5 YR 7/6 fine sand still withsome carbonate. Slightly lighter at 170cmfrom surface. Carbonate disappeared at 196cmfrom surface. Merging over 13cm to
53cm Reddish-yellow 5 YR 7/8 fine sand13cm Reddish yellow 7-5 YR 7/6 more clayey fine
sand (colour and texture change, merging to
28cm Pink 7-5 YR 7/4 fine sand with solid carbonatenodules which stopped the auger at 290cm.
This section is interpreted as follows:
Soil 1 0-1 01m A horizon1-01-1 -95m B horizon with carbonate
Substrate 1-95-2 -48m Dune sandDisconformity
Soil 2 2-48-2-61 m Remainder of A horizon oforiginal terrain or top of Bhorizon.
2-61-2-90+m B horizon of original terrain.
Nodules from the lowest part of the above
auger hole (site approximately 142° 15'E, 34°
46'S) were radiocarbon dated by Dr. T. A.
20 EDMUND D. GILL
Rafter, New Zealand Institute of Nuclear
Sciences (Lab. no. R 2729/4) with the result
of 27,500 ± 70 yr B.P. S18C w. r. t. PDB— —35°/oo< This is comparable with a dating
of 27,800 obtained for similar material from
N. of the River Murray (to be discussed later).
If my interpretation of the above section be
correct, then this date is the maximal age for
commencement of dune building at this site.
Indeed, it is probably quite in excess of that
time, because the date is an average of the
many years required to grow such a nodule. For
the purpose of this argument the C14 date is
taken at its face value.
4. Berribee. Station Paleosols
Berribee Station is situated in the far NW.corner of Victoria on the S. side of the MurrayRiver. Two types of dunes occur there, viz. the
E-W. longitudinal dunes of the Malice Dune-
field, and channel-bordering dunes of the
Coonambidgal complex on anabranches of the
Murray River. In the air photo constituting
Plate 4, three oblique zones of different types
of country can be readily discerned, viz.:
a. Coonambidgal Country, consisting of the
present Murray River channel to the N-,
the anastomosing Lindsay River to the S.,
other anabranches and past channels, all
with their respective iloodplains.
b. E-W. Dune Country occupying the S-W.sector of the photo. Zones 1 and 2 meet
in the NW. sector of the photo, where they
are separated by a cliff.
c. Relict Floodplain. This intermediate zone
in the middle and SE. sectors of the photo
is above present river floodings but has ap-
parently been subject to inundation in the
past. One use is for crops (as can be seen
in the photo) and is partly irrigated. Fromthe photo it can be seen that the E-W.dunes penetrate it. This is obvious on the
ground, as when the auger was sunk, it wasconsidered to be in the zone of the dunes.
It is thus an area of interdigitation of the
dune-building and riverine processes. Thesite is marked on the air photo in PL 4, and
is approximately 142°4' E and 34°8'. It is
about 2 6 km SSW. of the Berribee home-
stead (Mildura military map c. 408,785).
The site is on the summit of an E-W. dune,
and the auger penetrated:
38 cm Red sand 10 R 4/6 (moist)
190 cm Light red sand 2-5 YR 6/6 with earthy
carbonate fracbulcs up to l-3cm diameter
but generally smaller. Most can be broken
in the fingers. From 203-365cm from the
surface there is a greater concentration of
carbonates with fracbules up to 2 -5cm in
diameter. Carbonate zone 1.
36 cm Red sand 2-5 YR 5/8. No carbonate present,
hard lumps being due to cementation by clay,
pyrolusite and such.
53 cm Red sand 2-5 YR 5/6 with earthy carbonate
sometimes forming fraebules that can be
broken by the fingers. Carbonate zone 2.
132 cm Red sand 2 5 YR 5/8 without carbonate.
Lumpy as 36cm band above.
104 cm Gray 5 YR 6/1 line sandy clay with white
earthy carbonate and pyrolusite. Carbonatezone 3.
13 +cm Same matrix without carbonate.
This section is shown diagrammatically in
Fig. 10 to visually communicate that more than
half the profile is full of earthy carbonate. Myinterpretation of this section is:
i.w Dune
WUORIM N FORMATION
( arboaaie
Auger Hole
Rod S.md
LOVEDAY HEI>Ol)LRM
[•Wiiht land SuttjUX-
Gray Quy
'BLANCHE l"OWN CLAY
Fig. 10—Section of E-W dune c. 26 km SW ofBerribee Station homestead, S. of the RiverRoad, NW. Victoria (PI. 4). Site 19.
Soil 1 0- (M)-38 in A horizon of pedocal38-2-28 m B horizon
Soil 2 2-28-2-64 m A horizon2-64-3- 17 m B horizon
Dune Sand 3-17-4-49 mDisconformity at base of dune
Soil 3 4-49-5-53 m B horizon of truncated soil
profile on Pleistocene al-
luvial terraceAlluvium 5 53-5 66 +m Flood plain sediments.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 21
Other interpretations are possible. Thesaturation of this section with carbonate re-
quires explanation. There appears to be far too
much to be derived from the sediments them-selves. Not far away in S. Australia marine
limestone and later carbonate rocks are ex-
posed, which could have been a source of
windblown carbonate dust that was transported
by W. winds to the site. (cf. Chapman and
Grayson 1903, Walker and Costin 1971, Glasby
1971).
Dr. T. A. Rafter also assayed these car-
bonates for radiocarbon content and calculated
the following CI4 ages:
Carbonate zone 1 14,200 ± 790 yr B.P.
313C w.r.t. PDB = 3-5%oCarbonate zone 2 31,300 ± 1,200 yr. B.P.
S13C w.r.t. PDB = —2 l%oCarbonate zone 3 Beyond C14 range
lo-> 34,300 B.P.
2<r> 28,800 B.P.
813C w.r.t. PDB = -3-5%
For logs of two other auger holes in this
vicinity see section on sedimentology.
All the datings of the E-W. dunes are oncarbonate of the same type, and deposited under
the same conditions. Indeed much of it mayhave the same origin, blown in on the prevailing
W. winds that built the dunes. It is reasonable
therefore to directly compare these dates. Con-
sidering the variables in pedogenesis, the dates
match well. They indicate at least three periods
of terrain stability when soils formed:
3 — 16,000 yr. B.P.
2— 28,000 yr. B.P.
1 >C14After each phase of dune-building, the
morphology was depressed by weathering and
erosion, as is indicated by the near horizontal
soils.
The dunes of the Mallee Dunefield belong to
the Woorinen Formation, which will be dis-
cussed in the section on stratigraphy. Somegrain size analyses of the sediments will also
be reported later in the paper.
5. Lowan Sands
It has been mentioned that there is more
surficial instability of the terrain in the W. of
the Mallee Dunefield than in the E. This is
due to the tongues of Lowan Sands (Lawrence
1966, Fig. 8) that cross the area like giant
blowouts. The origin of these sands has been
discussed by Hills (1939), Crocker (1946)and Lawrence (1966). Because of the low
dynamics of the Murray River system the
Parilla Sand, Chowilla Sand, Diapur Sandstone
and other sand formations are all characterized
by unusually fine grades that make them rather
similar. It is therefore not easy to specify the
formations that have provided the quartz
sand for these dunes. These sands are poly-
cyclic. The sand forming the dunes on the E.
side of Lake Victoria was carried on site byriver action, and bears the general character of
the sands of this sedimentary basin. They are
thus not typical dune sands, but finer than usual.
Because polycyclic, they are well rounded and
well sorted, thus contrasting with the dunes
of the Simpson Desert (Folk 1971, p. 47)"but briefly removed from their poorly-sorted
fluvial source material".
6. Lindsay Island Dunes
This island is the area between the MurrayRiver and the anastomosing Lindsay River (an
anabranch). It consists of Coonambidgal sedi-
ments, and comprises mainly the following
types of deposit:
a. Greenish-gray to gray clayey sand flood
plain deposit (red gum flats).
b. Well washed light brownish gray channel
sands.
c. Channel-border dunes (with Murray pines)
developed by aeolian action on the channel
sands.
Channel border dune .it ilic (. < KJYWIMIX.AI. fURMATIONTussock*
Ncwlv Mown vind'
Fig. 11—Surveyed section of channel border duneoriented N-S. on Lindsay Island. BerribeeStation, NW. Victoria.
22 EDMUND D. GILL
The air photo constituting Plate 4 shows part
of this area, and in particular the two-mile long
N-S channel bordering dune which is part of
site 19. Figure 11 is a section across this dune,
and the logs of the auger holes there shown
are as follow:
Auger 1 99cm Light brown 7 5 YR 6/4 to light
yellowish brown 10 YR 6/4 silica
sand merging to
38cm Pale brown 10 YR 6/3 sand merg-ing to
137cm Light yellowish brown 10 YR 6/5sand
282cm Pale brown 10 YR 6/3 sand15cm Compact mottled light gray 10 YR
7/1 and reddish-yellow 7 5 YR 6/6clayey silt.
This profile is interpreted as:
0-5-56 m Dune sand with little variation
(oxidized).5 • 56-5 7 l
+m Riverine Coonambidgal sediments(reduced),
Auger 2 10cm Dusky red sand 2-5 YR 3/2 moist,
merging to
28cm Reddish brown 5 YR 4/4 sandmerging to
53cm Reddish yellow 7-5 YR 6/6 sandmerging to
41cm Yellowish red 5 YR 5/6 sand merg-ing to
79cm Light yellowish brown 10 YR 6/4sand merging to
107cm Very pale brown 10 YR 7/3 sand58cm Mottled clayey fine sand—light gray
near 5 YR 7/1 and reddish yellow7-5 YR 6/6. Occasional streaks area stronger red. At 3 07m carbonpresent and iron oxide straight nar-row markings that appear to be rootsites.
137cm Similar sediments with changingpercentage of fines and changingamount of mottling.
This profile is interpreted as:
0^3 -71m Oxidized aeolian sand with a thin
juvenile soil.
3 • 7 1—5 * 36m Gleyed riverine Coonambidgal sedi-
ments
Auger 3 10cm Pinkish gray 7-5 YR 6/2 well sortedand well-washed sand.
36cm Light gray 10 YR 7/1 poorly sortedsand including grades heavier thanin dunes
33cm Lighter gray N 7 sand, poorlysorted and with clay balls. Augerbrought up half of one 3 -8cm indiameter plus other pieces. Mergingto
8cm Light brownish gray 2-5 YR 6/2sand with a few streaks of light olive
brown 2-5 YR 5/4.28cm Offwhite (near N8) sand
The final 36 cm of the hole were difficult to
drill as the sand was running so freely. The log
is interpreted thus:
0^ 10cm Dune slope wash10-175cm Oxidized, mostly well-washed riverine
sands less sorted than dune sand derived
from them.
Grain size analyses have been carried out on
some samples, and are reported later in this
paper.
Cal Lai Dunefield
In the study area a distinct difference exists
between the country N. of the Murray River
and that S. of it, so that the course of the
Murray may be a function of this difference,
modified by the minor tectonics already des-
cribed. The Mallee Dunefield to the S. is a
complex of E-W. longitudinal dunes with
pedocals of earthy carbonate. It is a sandy
terrain. The area N. of the river is characterized
by a substrate of impermeable green Blanche-
town Clay with a complex pedocal containing
solid crystalline calcite nodules. At first the
terrain appears to be a typical sandy one, but
closer examination shows that the sand is
largely a veneer. The excavations for water
"tanks" show this well, as also did our spade
and auger holes. The air photos show that someareas N. of the river have well-developed dunes,
and some have none, but overall an E-W.grain is present, developed undoubtedly by the
persistent W. winds. Dunes N. of the river are
best developed in the study area close to
1 \\ [June SjsWtii
WOORIMEN FORMATION Suifici.il sum! oriented E-W
RLTl/S I ( >R MAT10V^\~:- -~\Z£.~ ~,
RIVER MURRAY VALLEY
Predominantly Clau-v"
COONAMBIDGAL /—
~
FORMATION
Fig. 12—Notional section of Mallee and Lake Vic-toria dunefields to show their differentcharacter due to different substrates. Theriver follows the interface. Thus farmers inthe Mallee have difficulty retaining irriga-tion water, while N. of the river water tanksare dug directly into the surface.
VK HtKIA
Fig. 13—Map of Lake Victoria and environi ihowing features pertinent to this study. Based
on Engineering and Water Supply Chowilla Dam map (S. Australian Government),
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 23
sources of sand e.g. E. of Lake Victoria. Thedifference between the country N. and S. of
the river is fundamentally that the S. country
has a sandy substrate (Diapur sandstone,
Chowilla Sand and such) while the N. country
has a clayey substrate.
It is to be expected that the river would find
and follow the interface between the two (Fig.
12). In the short time available for this Project
the idea could not be developed, but its clarifica-
tion seems important for understanding the
country and for the planning of irrigation, water
conservation, groundwater utilization and such
like.
Mt. Hancock Dune
Any eminence in a flat terrain appears im-
portant. From considerable distances one can
see a large dune of mobile sand that has been
given the name of Mt. Hancock. It surmounts
the high cliffs on the E. side of Salt Creek near
its confluence with the Murray River, and is
approximately N. of Kulcurna Homestead. The
structure is conspicuous because so elevated,
isolated, and of light colour due to absence of
vegetation. Its origin is a function of lateral
differentiation and placement relative to prevail-
ing winds. At this cliff site, the Blanchetown
Clay cuts out, and the Chowilla Sand takes over
a great part of the cliff section, as can be seen
also on the right bank of the Murray River be-
tween Kulcurna Homestead and Salt Creek. Asthe Chowilla Sand is a well-washed sand, it
breaks down to loose sand very readily ex-
cept where it is affected by the paleosol that
mobilised silica. The cliff below Mt. Hancock
has thus a good supply of free sand, an adequate
source of material for dune building. As the
cliff is on the E. bank of Salt Creek, the pre-
vailing W. winds supply the necessary traction.
As W. of the creek is a very extensive lowland
flood plain, the winds are able to exercise their
full force.
Lake Victoria
"Beneath me lay a beautiful lake, about 30 or
40 miles in circumference, with a line of gum
trees scattered along its edge ... To this
splendid lake of which I had the pleasure of
being the first European discoverer I gave the
name of Lake Victoria" (Hawdon 1852, pp.
42-3). Hawdon made his journey in 1838
and named the lake after Queen Victoria.
Shortly afterwards Sturt (1849, vol. 1: 94)gave this description
—"Lake Victoria is a very
pretty sheet of water, 24 miles in diameter, very
shallow, and at times nearly dry. It is con-
nected with the Murray by the Rufus, and by
this distribution of its waters the floods of the
Murray are prevented from being excessive".
It should be noted that there was some early
confusion in nomenclature. For example Bon-wick (1855) refers to Lake Victoria but his
maps show two lakes of this name, viz. (1).
What is now called Lake Alexandrina at the
seaward terminus of the Murray River, and (2)
Lake Victoria, between the River Darling and
the S. Australian border.
"Among the unlicensed Murray squatters
were Edward Meade Bagot . . . and George
Melrose who came up the river in 1847 to
occupy the bush frontages of Lake Victoria.
The latter's shepherded flocks did so well that
he would have liked to legalize his claim but
this was impossible. Until the position of the
border was fixed, he discovered, neither colony
was in a position to grant a lease. The charms
of Lake Victoria so attracted Melrose that he
chose it to honeymoon with his Scottish bride,
Euphemia—as the first white man's lubra to be
seen in those parts she caused quite a stir
amongst the Aboriginal tribes" (Hardy 1969,
p 63). Melrose, tired of waiting for a lease,
went elsewhere. The surveys were not com-pleted until 1854. However, we have here a
picture of a beautiful lake with vegetated banks
when first visited.
The foregoing descriptions will again be rele-
vant when present day erosion patterns are dis-
cussed.
Lake Victoria (Figs. 4, 13) is 13 km long
and 10 km wide with an axis oriented a little
W. of N. Under natural conditions only the S.
end was full, a heavy flood being needed to
cover the whole present lake bed. In the early
days of settlement the Wentworth to Renmarkroad or track crossed the N. end of the lake
bed as also did the telephone line. Lake Victoria
24 EDMUND D. GILL
(PI. 5) is now a reservoir. Waters held in the
Menindee Lakes on the Darling River can bereleased down that stream and by control at
Lock 8 diverted through Frenchmans Creek(bordered by artificial levees) into Lake Vic-
toria. They are released through a regulator
down the Rufus River and so into the MurrayRiver again. Although 732 km by river fromthe sea, Lock 8 is only 25 m above sealevel.
Distinct from the existing Lake Victoria is
the basin of which the present lake occupies
only a part. This basin is sunk into the plateau
of Blanchetown Clay and associated formations
so that the level of the reservoir is 30-35 mbelow plateau level. The water is about 16 mmaximum depth and an unknown depth of
sediments lies below that, but these are shallow
judging by the mode of genesis of the lake. Thebasin thus probably reaches a maximum of the
order of 38 m below the general plateau level.
(The present lake is pressed against the W.perimeter of the basin, which is scalloped by a
series of erosion amphitheatres (Fig. 14) that
are separated by promontories somewhat lower
than the plateau level. On the E. shore of the
lake is a large lunette, about 14 km long and upto 2-4 km wide. To the NE. of the lake is a large
area approaching half the size of the lake con-
sisting of lunette backed by swamps, claypans
and relics of Rufus Formation. Separated by a
promontory from this NE. area is one to the
SE. which is of different character again, con-sisting of ancient Coonambidgal floodplain andchannel sediments through which stand outliers
of Rufus Formation. This area stretches SE.
to Moorna Station and S. to Frenchmans Creek.
Thus in the Lake Victoria basin there are four
geomorphic elements:
1
.
The erosion amphitheatres on the W.perimeter.
2. The lake.
NOOl \ M VI h IN
3.
4.
The NE. area with lunette, swamps, clay-
pans and Rufus Formation outliers.
The SE. area composed of greenish grey
Coonambidgal Formation sediments with
higher red relics of Rufus Formation.
On a map the lake is seen to be a smooth-
shored oval stretch of water but the basin in
Fig. 14—Erosion amphitheatres on the W. shore ofLake Victoria, N.S.W. Based on air photos.Site 6.
which it lies has no such character. The basin is
about twice the area of the lake, but in view of
its shape, it is not easy to decide where its S.
boundary should be. The perimeter of the
basin is irregular on the W. because scalloped,
and in the E. has broad rounded sweeps. Anyexplanation of the genesis of the lake mustaccount for these features.
Genesis of Lake Victoria
If the Lake Victoria basin is consideredapart from the present lake and lunette, it is
seen to be an area wide open to the S. It is notenclosed with a narrow outlet as is the presentlake due to the encircling W. perimeter of the
basin on one side and the lunette on the other.
The axis of the basin is similar to that of the
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RTVER 25
lake, i.e. deflected to the W. As the Blanche-
town Clay and associated formations continue
round the walls of the basin without appreciable
change, the basin does not owe its presence to
faulting. It is erosional, scoured by the an-
cestral Murray (or a large anabranch, or both)
deflected NW. from the present course. It
would appear that the stream left the present
valley tract in the general area of MoornaStation homestead (Parish of Moorna) and
ran more or less NW. through the Parishes of
Utah and Wangumma to the Lake Victoria
region. The Lake Victoria basin is thus re-
garded as a giant billabong and the scalloped
margins are probably remnants of the Pleisto-
cene meanders there. The time of formation
was after the deposition of the Blanchetown
Clay (Plio-Pleistocene) because, with as-
sociated deposits it forms the plateau. On the
other hand, the basin was cut before the deposi-
tion of the red Rufus Formation because relics
of this terrace occur in the main Murray River
tract, in the open S. end of the basin (Dunedin
Park Station) and in the more restricted NW.area of the basin behind the lunette. After the
Rufus Formation was formed, some change in
the river regime resulted in cessation of terrace-
building, deeper scouring of the river tract, and
then deposition of the Monoman and Coonam-bidgal Formations. This explanation of the
Lake Victoria basin accounts for its strati-
graphy and geomorphology. It also accounts
for its W. trending axis. That the lake is pressed
against the W. perimeter is consonant with this
theory. Sand carried in by the river was blown
up over many thousands of years by the per-
sistent W. winds of the area to form the lunette
on the E. side of Lake Victoria. Its position is a
balancing of the water pressure against the W.
perimeter and against the accumulating sand
on the E. side. With the diminution of river
flow into the basin, the lunette extended S. and
cut off the water supply or forced the inlet
channel S. so that water flowed only through
the Frenchmans Cr. anabranch. Thus the pre-
sent lake was established.
The genesis of the lake is far more com-
plicated than this simplified account. The
lunette had numerous alternating unstable
(dune-building) and stable (soil-forming)
phases. These can be divided into two periods
represented by the Nulla Nulla and Talgarry
Members of the Lake Victoria Sands which
constitute the lunette (see section on strati-
graphy). Some deep valleys were formed and
a general depression of the dune morphology
effected between emplacing the two Members-
There was a marked change of fauna because
large extinct marsupials characterize the Nulla
Nulla Member while the modern fauna only
occurs in the Talgarry Member.The Lake Victoria lunette (for term see
Hills 1940, discussed Gill 1964, p. 353, Camp-bell 1968, Macumber 1970) has an unusual
morphology. More typical ones occur at LakeWallawalla, S. of Lindsay Is. (Fig. 15), and
Fletcher Lake, N. of Curlwaa.
The unusual character of the lunette lies in its
great width and commonly horizontal bedding,
which are related factors. Typical dune mor-phology from sand blowing up windwardslopes, and spilling down lee slopes does occur,
but more often the bedding is flat or near it.
Similar structures occur in modern blowouts
where sand is blown horizontally for great dis-
tances, then slides down a normal sandfall. It
is therefore inferred that the supply of sand wasinadequate in relation to the energy of the
W. winds that provided the traction. Also as
the sand was so fine, it was more easily moved.The periods of dune stabilization could there-
fore be due to:
1. Cut-off of sand supply (e.g. by change of
river course),
2. Change in wind regime (considered un-
likely) or
3. Increase in precipitation causing a stronger
vegetative cover.
The factor yet to be explained in this theory
of the genesis of Lake Victoria is what caused
the river to divert to the NW. Attention is
drawn to the large fault at depth below the E.side of Lake Victoria described in the section
on tectonics. In this terrain of such exception-
ally low declivities, a small movement on this
fault could cause the river to be diverted. Asthe diversion occurred immediately after the
26 EDMUND D. GILL
VICTORIA
SITE IK
Nl'J's Corner Sin.
Ned's Corner Stn.
Fig. 15- -Lake Wallawalla and an unnamed similar dry lake to the W. of it, both with lunettes.
The age of these structures has not yet been determined. Based on Engineering andWater Supply Dept. (S.A.) Chowilla Dam map.
deposition of the Blanchetown Clay (i.e. just
after the draining of Lake Bungunnia) a nearhorizontal lake-floor terrain would be present
where the smallest tectonic disturbance wouldhave a positive influence on the direction of
river flow.
Lake Victoria Gulches
In this semi-arid country (22-25 cm rain-
fall) where rain seldom falls but may fall in
quantity over a short period, a characteristic
landform is the washout or gulch—short, steep
and often deep valleys tending to vertical sides.
It is comparable with the wadi of the MiddleEast, the arroyo of certain parts of NorthAmerica, and such. The term "gully" is often
used in Australia, but usually refers to an open,
short valley in humid areas, where vertical sides
are not usually found. Thus the term gulch has
been chosen, and is used herein as defined
above. Its derivation is from a Middle English
word that meant to "swallow or devour greed-
ily". Thus the Shorter Oxford Dictionary states
that the word means "a narrow and deep ravine
with steep sides marking the course of a tor-
rent". Twice during our field work one third of
the year's rainfall fell in one storm. Heavyrain in country of poor vegetative cover causes
brief gushing torrents that rapidly corrade. This
is particularly so since European settlement,
which has often been accompanied by too
heavy stocking. The first settlers came froma humid country, and did not understand this
dry land.
Lunette Gulches
The most notable system of gulches in the
area of study is that along the E. shore of LakeVictoria where the dune front has been severely
eroded. These numerous gulches have providedimportant stratigraphic, pedologic, sedimento-logic, archaeologic and palaeontologic data, andmerit detailed study. Monash University De-partment of Zoology under Professor J. W.
NULLA NULLA S1A1 [ON
N3 SITE ()
\2 Gulches
SIT! S
TALGARRY STATION
NEW SOUTH WALES
lo km
Fie 16—Map prepared from air photos of the gulch system in the lunette on the E. side
of Lake Victoria, N.S.W. Fixing of fossil and other sites was difficult, so the gulches
have been numbered N. and S. of the Nulla/Talgarry fence, the only permanent
marker. Numbered pegs have been set at the heads of most of the gulches. Four
deep concrete bench marks were inserted for fixing levels. As the lunette structures
are subhorizontal (on the whole), vertical levels have significance.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 27
Warren is continuing the study of the verte-
brate fossils from this and other localities. Withrecurrent gulches of similar morphology,changing somewhat from year to year, wefound it rather difficult to record sites so that
they could be located again with speed andaccuracy. It was therefore decided to numberthe gulches. As the only permanent mapped fea-
ture is the boundary fence between Nulla andTalgarry Stations, the gulches were numberedN. and S. of this line with appropriate N. or
S. prefixes (Fig. 16). Over most of the lunette,
square hardwood pegs 12m long and with
sides 5 cm wide were inserted at the gulch
heads and the relevant number drilled by a
series of holes into the leeward side. After twoyears they show no sign of deterioration. Fourconcrete benchmarks were also emplaced, and
surveyed on to Chowilla Dam Survey pegs so
that levels can be related to Murray River and
sea level datums.
The morphology of the lake front is a wedge
of sand blown up from the lake beach, modified
by fans of outwash sand from the gulches. Themorphology of the gulches varies. Theephemeral streams corrade steeply, tending to
develop vertical walls. Wind erosion cuts deep
blowouts with steep sides. This geometry is
modified by rainwash, aeolian action and the
activities of sheep, goats, rabbits and man. Thereduced stability since settlement has madethe area rather barren in contrast with the
natural condition described by the discoverers.
The hamada of outwash fans along the lunette
above the beach appears to be recent. The sand
slope was measured at 7°. About the middle of
the E. shore (air photo CAC 37-5012 Lake
Victoria run 5, 1954, 23 mm S. of E-W. line
through CP) an auger hole was sunk 10 mfrom the water's edge on 25th Nov. 1969.
3 m Very pale brown 10 YR 7/3 (dry mediumloose sand very poorly sorted, grading to
0-45 m Brown 7-5 YR 5/4 (moist) ditto grading to
0-6 m Damp ditto grading to0-45 +m Dark brown 7-5 YR 3/2 (wet) ditto.
The sand was uncompacted and water fromthe lake penetrated to the bore hole makingfurther boring impossible with an auger. It
was hoped to reach a carbonaceous former lake
bed deposit that could be dated by radio-
carbon.
(b) West Bank Gulches
The lunette gulches are mostly single chan-
nels (Fig. 16), 65% being so classifiable. An-other 24% are simply branched, while the
remaining 11% have multiple branches. It is a
matter of interpretation which gulches go into
which category, but these figures are of the
correct order, and indicate a simple pattern of
gulch formation. The erosion pattern on the
W. bank is quite distinct, and the difference is
a function of substrate. The lunette gulches
are in fine sand with some clay, while the W.bank ones are eroded mostly in Blanchetown
Clay and Chowilla Sand. From plateau level
there is a steep slope from 10° to 30° then
a lower slope to the lake edge.
Fig. 14 shows a tracing from air photos of
these amphitheatres and the dendritic pattern
of the gulches. Fig. 13 also shows a similar
development on the N. bank of the lake W.of the "Old Hotel" (the site of an hotel be-
longing to the coaching days when this was
a stopping place between Wentworth and
Renmark).
Peg on plateau edge NOOLA EROSION AMPHITMH VI RE
r-i^S*. ni.ANCMIiTOWNCLAY
Fi&h FossHsFormer lake level?
k Gulch I,
£ .'.'.'.'. Horizontal Scale *
. <
> • '
.
''
.
"'
.
' .100 nV.
CHOWILLA SAND" Former lake level'
Lake Victoria level when lull
Fig. 17—Surveyed section of the Noola Erosion Amphitheatre, NW.shore of Lake Victoria, N.S.W. Site 6.
28 EDMUND D. GILL
(c) River System Gulches
Gulches occur along the river banks(Murray, Darling and Anabranch ) , roundsome billabongs, and along creeks, such as Salt
Creek and Six Mile Creek. Some of the best
developed are on the E. side of Salt Creek onKulcurna Station. One had such bright oxida-
tion colours that we called it Kaleidoscope
Gulch. One gulch N. of this was characterized
by a cirque-like head. In this area, the silicified
Chowilla Sand forms a lower plateau level as
at Cal Lai. This zone is resistant to erosion
so deeper gulches with steeper sides result.
Pieces of the silicified sand block the gulch
floor in places. This rock contains rhizomorphsup to 5 cm diameter. The indurated zone formsa shallow syncline, which causes variation in the
depth of the gulches.
Gulches round Triple Swamp have yielded
dolomite, fossil bones and Aboriginal sites
as well as extensive stratigraphic information
as recorded later in this paper. Much of the
information discovered is due to the develop-
ment of this feature. As every heavy rain
gouges them out a little more, fresh sections are
always available.
Stratigraphy
Geologists "first tracked mud andheresy into the schools of learning
200 years ago". Fuller (1971)
1. Standard section
In connection with the proposed ChowillaDam, bores were sunk across the Murray Rivervalley at a site between Renmark, S.A., andthe Victorian border (Figs. 1, 4). J. B. Firmanstudied the geology and prepared a report whichhas not been released. However, a summary of
the stratigraphy has been published (Firman1966, 1967). Firman kindly provided a copyof his cross-section (Reproduced in Fig. 18)and this well-attested profile is here used as a
standard section. For the present purpose, the
following are noted:
( a ) Superposition
The succession is simple, so the relative ages
are clear.
(b) Tectonism
After the Parilla Sand was deposited, earth
movements belonging to the Kosciusko Epochcaused uplift of the Pinnaroo Block (=MurrayRidge of earlier writers), damming the Ter-
tiary river drainage, and so causing deposition
of the Blanchetown Clay. Later, penetration
of this natural barrage led to downcutting andexcavation of the broad (4 8 km) gorge (so
called because of depth and steep to vertical
banks) into the Blanchetown Clay, ChowillaSand, Parilla Sand, and Loxton Sands.
A disconformity thus exists between the
Tertiary Loxton Sands and the Late QuaternaryMonoman and Coonambidgal Formations.
(c) Palaeoecology
The Murray Basin has a succession of
Tertiary marine strata (Fig. 3), the latest of
which in our study area is the BookpurnongFormation, which was also encountered in a
well excavation (Tate 1 899 ) on TareenaStation (Fig. 4), and in all the oil bores (Fig.
3). The succession of facies with time (i.e. up-wards through the section) portrays the re-
treat of the sea from the Murray Basin, viz:
Parilla Sand (clayey)
Chowilla Sand (well
washed)
Blanchetown ClayLoxton SandBookpurnong
Formation
Pata Limestone
Riverine floodplain
Riverine channel
Lacustrine
Estuarine
Marine (shallow
water)
Marine (shelf)
With this principle of facial succession estab-lished, one would expect to find a similar suc-cession throughout our limited area of study,
and this appears to be so. The A.O.G. Went-worth No. 1 Bore (40 km N. of that town) re-
vealed Loxton Sands or equivalent at 52*5-80- 5 m and Bookpurnong Formation belowthat to 109-1 m.
The A. O. G. Lake Victoria No. 1 Bore (at
the S. end of the lake) penetrated "typical"Bookpurnong Formation from 93 9-1 13 -4 m"with a rich microfauna". Above this are"pyritic sands with carbonized wood fragmentsand fish bones", which might be estuarine and
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 29
the equivalent of the Loxton Sands. They aresimilar to the sediments in the Wentworth Boreat that stratigraphic level. The A. O. G. NorthRenmark No. 1 Bore (about 8 km N. of that
town, near the corner of the Ral Ral Ave. andthe Wentworth Road ) encountered LoxtonSands from 18*3-42-7 m. Bookpurnong sedi-
ments were reported from 42-7-61 m. Thus in
all three bores the Bookpurnong Formation is
adequately defined but the Loxton Sands are
proven only in the Renmark Bore. Even there,
definitive fossils (small foraminifera, echinoid
spines and molluscan fragments) occur only in
the lowest 3 m. So perhaps the sediments in the
other bores are riverine equivalents of the
estuarine Loxton Sands.
(d) Structure
The strata (including the non-marine) in the
standard section are superposed in a regular
manner with subparallel surfaces. This regular-
ity does not continue in the non-marine forma-
tions of the study area to the east. For example,
the Chowilla Sand does not maintain the fixed
stratigraphic position it possesses on the Pin-
naroo Block. On the contrary, it appears below
the Blanchetown Clay, within it, or above it, or
any combination of those. Thus E. of the Pin-
naroo Block it is a facies indicator, and nolonger a stratigraphic marker.
Structure E. of the Chowilla Dam site can be
defined by the highest indubitably determined
stratigraphic horizon in the A.O.G. section
(Fig. 3), viz. the base of the BookpurnongFormation. Adjusting the levels to the sealevel
datum, the results are:
37*4 m83-7 m69-5 m
Renmark Bore
Lake Victoria BoreWentworth Bore
In compensation for the uplift of the Pin-
naroo Block, the strata to the east sagged, and
the resultant structure is here called the Lake
Victoria Syncline. As this syncline was due to
upthrust of the crustal block to the W., the mild
fold is understandably asymmetric with a lower
dip on the E. flank. The openness of this syn-
cline is in keeping with the minitectonics (for
term see Gill 1972) of the area. The Lake
Victoria syncline explains why:
L The Parilla Sand, although a well charac-
terised formation disappears below river
level east of Kulcurna Station. As far as I
have been able to check the stratigraphy, it
does not reappear until the Paringi area
29kmSE. of Mildura.
II. The characteristic interdigitation of the
formations upstream from the dam site,
i.e. within the syncline. The Pinnaroo Blockconstituted a natural dam, and created a
temporary or structural base level (Chorley
and Beckinsalc 1968). However, the gross
interdigitation means that the lake so
formed was not one large sheet of water
that persisted over the whole area through-
out the period of impedence to river flow,
but rather a fluctuating series of lakes andswamps across which river channels kept
changing. The high evaporation no doubthad much to do with this. Widespread bedsof dolomite developed at one time, and in
some areas these were eroded before de-
position continued, while in others the very
fine dolomite was mixed with sand. Aswould be expected, the most continuous
lacustrine conditions prevailed near the
damming Pinnaroo Block. Thus the thickest
sequence of Blanchetown Clay can be seen
along the Murray River at Lake Victoria
(Warrakoo) and Nampoo Stations (Fig.
4).
2. Regional Formations
The formations consist almost entirely of lowdynamics sediments, comprising clays to fine
sands. This applies to both the aeolian andriverine formations, because the former are de-
rived from the latter. The coarse fractions of the
originating streams are trapped in the E. side of
the Murray Basin because the traction was not
available in this flat country to bring themfurther W. Significant in the stratigraphy there-
fore are:
(a) The presence of any gravels, as they are
so rare.
(b) The change from sand to clay, and vice
versa.
(c) The change from clayey sand (floodplain)
to well-washed sand (channel) deposits.
30 EDMUND D. GILL
(d) The presence of dolomite and limestone,
because they indicate special sedimentary
conditions.
(e) The presence of paleosols because they
indicate a cessation of sedimentation, a
time-break, and an absence of water or
other such cover. They indicate also a
period of stability, and the extent to which
the subaerial agencies could leach (and
otherwise alter) the surface of the terrain.
The nature of the minerals removed or
mobilised indicates the prevailing geo-
chemical conditions, which have implica-
tions with respect to the vegetative cover
and fauna.
The paleosols are described in an accompanying
paper on palaeopedology.
The formations will now be briefly described
from oldest to youngest.
(a) Bookpurnong Formation (Figs. 3, 18).
This is the youngest of the marine formations;
it extends throughout the studied area. It is
therefore used as a stratigraphic and chrono-
logic datum. Questions of chronology will be
dealt with in the section on that subject. Al-
though it does not outcrop at the surface it was
G7- -G36I r PD59(D400)
PD86
PDI
PD8I kuzu^MJRRAf P'VfK
PD20-I PDI2-
'
~^
VERT.
SCALE
METRES 5000
FEET 2000
SCALE10000 METRES
6000 FEET
MARINE
TT7^
ESTUARINE
LEGENDFLUVIAL-LACUSTRINE
FLUVIAL AEOLIAN SOILS
Upper Villey hit:
[ undifferentiated]
Lower VaU,= , F !.
"oncmin Formaupn
Dcpows of
high
level rrte n :•
!
'-—^ Blanchctown Clay
— Siliof.cd Cap on Parilia Sand
Clay Bed -Par, Hi Sand
Loxton Sands (Upper 1 Bed)
Loxton Sands (Lower Bfd r
-)
•y*- j J -j Bookpurnong Beds
~TT
LSiis Rata Limestone equivalent?
Woorintn Formation
(Right embankment)
S A Deot. of Mir,
Fig. 18—Standard stratigraphic section at the proposed Chowilla Dam site, by courtesy of the
S.A. Dept. Mines. Site 1.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 31
encountered in a well on Tareena Station and
can be followed in the three A.O.G. bores. For
lithology and palaeontology see Ludbrook et
al. 1958, Ludbrook 1961, 1969, Firman 1966,
1967.
(b) Loxton Sands (Figs. 3, 18). This forma-
tion of fossiliferous estuarine sands forms the
cliffs at Loxton, S.A., is present at the base of
the Chowilla Dam section (Fig. 18), occurs in
the Renmark bore (Fig. 3), but appears to
grade into a riverine facies further E.
The succeeding formations occur in the Lake
Victoria syncline in river and lake bank out-
crops:
(c) Parilla Sand (Figs. 4, 18-20). The gray
compact clayey sand, with a vertical cleavage
is very consistent and easily recognized. Where
cliffs are undercut in river or gulch it forms
vertical faces, and where not it forms steep
rilled slopes. The section on the right bank of
the River Murray at Kulcurna Station just
downstream from the homestead is unusual in
that it consists of only two formations—Parilla
Sand surmounted by Chowilla Sand (Fig. 19).
A similar thickness occurs on the S. bank of the
River Murray at Boundary Point (Victoria/SA.
boundary). The cliff was measured nearest the
homestead where it is lower but more acces-
sible, viz.:
Exhumed Karoomla Terrain
5 m Yellow 2-5Y 76 to Reddish yellow 7 5\ 1< 7 6
Silicified Columnar CHOWILLA SAND
KAROONDA PEDODERM
2-2 m While 10YR 8/1 with reddish veltcw jnd' • - •"•_ *
weak red 7 5R 4/4 mottlesJj_.
'
___-==< 33"
ni ( .n ,_M,->
033 m Molded Clayey Sun-'
5 • m Grav Compact
PARILLA CLAYEY SAM)
Murray River Lc
Fie 19 Measured section of cliffs beside Kulcurna
Station homestead, Cal Lai, N.S.W. Right
bank of Murray River. Site 3.
5m Yellow (2-5 YR 7/6. and variants) to red-
dish yellow (7-5 YR 7/6) lightly silicified
columnar sandstone forming vertical cliff
Chowilla Sand2- 2m Mostly white (10 YR 8/1) silicified sand-
stone with reddish yellow and weak red
(7-5 R 4/4) mottles; rhizomorphs, andsilicified burrows. The top of this bed is moresilicified than elsewhere, and forms the
upper hard band seen in the cliffs. Similar
horizons occur at Salt Cr. and the Chowilla
Dam site—Chowilla Sand.
0-33m Black to gray clay. Causes change in cliff
slope. Recognizable on inaccessible faces of
the cliff by course cracking. Minute crystals
of gypsum on some joint planes
33m Mottled gray, light brown and red clayey
sand with fine cracking.
015m Plaley silcrete in one to six layers. Material
suitable in places for Aboriginal implements.This forms the lower hard layer seen roundthe cliffs
5 + m Compact gray (10 YR 5/1) clayey sand
—
Parilla Sand.
13 01m
The thick lightly silicified columnar sand re-
minded the author of a similar deposit above
kaolin in the Home Rule pit near Gulgong,
N.S.W. The white sand at the base of the
Chowilla Sand varies in thickness. The forma-
tion is an aquifer while the underlying Parilla
clayey sand is an aquiclude, so the whiteness
may be a function of leaching along the base of
the aquifer. There is a sharp boundary to the
underlying clay, so the ancient river spread sand
over a lagoon or swamp. The clayey sands are
interpreted as a floodplain deposit and the well-
washed sands above as a channel deposit Theclean sands readily admitted air, and so are
oxidized.
The degree of silicification varies greatly. It
is less near the mouth of Salt Creek so that loose
sand has blown up to form the eminence called
Mt. Hancock. At the N. end of the Salt Creek
cliff is silcrete, that when free to vibrate rings
like a bell if struck.
At the outlet of Salt Creek and for a short
distance upstream the section is essentially as
above. Then the Blanchetown Clay comes in,
and a measured section marked by a peg at the
S. end of Scrub Paddock is as shown in Fig. 20,
viz:
4 1m Red Bunyip (?) Sand21m Red and yellow silicified Chowilla Sand1 2m White ditto
32 EDMUND D. GILL
4 m I-:IJ) SAM)
K.mmmihLi 1K \|«.<)N|>A ]<!])( >!>bRM2 I ni Red and yellow silinlial
< MOWII.LA SAND
Shiij p boundur)
1
3 m BLAW HETOWN CI AY
Sharp boundary
">i How and v. lii
J
' n PARTI LA CLAYEY SAND*"
I base ooi .t.-i-n)
Silll ( r,vk k-\c\
Fig. 20—Measured section on F. bank of SaltCreek, Kuleiirna Station, N.S.W., at
the S. end of Scrub Paddock. Site 2.
2 3m Green Blanchctown Clay3-0m-f Gray compacted clayey Parilla Sand 8 3m
to creek level covered by talus
Firman (1966, 1967a,b, 1971) gives theParilla Sand as the equivalent, at least in part,
of the marine Norwcst Bend Formation. The1972 S.A. Department of Mines map of sur-
ficial deposits in the Murray Basin shows asuccession of ecologies in the formations as theyappear from W. to E. on the Murray River fromMorgan to the Victorian border, viz.:
Morgan Limestone (marine shelf)—LoxtonSands (Estuarine)—Parilla Sand (Riverine)—Blanehetown Clay (lacustrine).
The formations are thus younger from W. to
E., and pass from marine to terrestrial fromW. to E. In the study area the Parilla Sand is
overlain (with a sharp break) by either
Blanchctown Clay or Chowilla Sand. In someplaces there is evidence of slight erosion on thetop of the Parilla Sand, but in general the topof the formation is remarkably horizontal.
The Murray River cliffs between Paringaand Loxton, S.A., show excellent outcrops ofParilla Sand. For example, where the Sturt
Highway meets the cliffs 2 4 km NW. of the
Loxton turnoff (12 km from Renmark) near a
parking bay, the Parilla Sand constitutes ap-proximately the lower half of the cliff (PI. 2fig. I, PI. 6, fig. I). The Parilla Sand can beseen to continue for some kilometres along thecliffs, but it cuts out E. of Salt Cr. (Kulcurna
Stn.). PI. 8, fig. 2 shows the occurrence there
with its typical vertical cleavage and rills.
Parilla Sand (Fig. 18) is the main cliff out-
crop on both banks of the Murray River at the
Chowilla Dam site (PL 2, fig. 2). The top of the
formation is marked by a bench consisting of
the silicified materials of a palcosol (see paper
on palaeopedology). The top of this fossil soil
is the Karoonda Surface (Firman 1967) or
Terrain.
The formation is named after the town of
Parilla in S. Australia (Firman 1971 and re-
ferences). Firman (1971, p. 62) claimed that
the Parilla Sand could be traced "as far up-stream as Nyah in Victoria". In this MemoirParilla Sand is retained in its original definition,
and the very varied sediments lying immediatelybelow the Blanchctown Clay (except ChowillaSand) in the Lake Victoria Syncline are des-
cribed as the Moorna Formation. Parilla Sandprobably occurs below this. The uplift of the
Pinnaroo Block occurred in post-Parilla time,
and earlier than the deposition of the Blanche-town Clay.
(d) Blanchctown Clay. Firman (1965, 1971)provides the origin of this stratigraphic nameand maps the extent of the formation in S.
Australia. The type section is on the left bank ofthe Murray River 4 8 km below Blanehetownin S. Australia. The formation is distributed
"North and east of the Pinnaroo Block" (Fir-
man 1971, p. 54). It may be described as themost characteristic formation of the study area.
On present knowledge, the thickest outcropsare in the River Murray cliffs on Warrakoo andNampoo Stations and on the banks of LakeVictoria, i.e. in the Lake Victoria syncline. Theformation thins out to the east (Fig. 36) butcontinues at least as far as Robinvale, whereBlanehetown Clay outcrops in the river cliffs.
The lithology is typically a greenish gray clay-stone to siltstone. A. J. Gaskin, C.S.I.R.O., sug-gested that the green colour is due to ferrousiron in the lattice. As may be expected, theclays grade in places to sandy phases. See Seg-nit, this Memoir.
Figure 21 is a measured section E. of Nam-poo Station homestead on the W. side of atrack down the cliff to river level. At this
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 33
13 an Hurdpan reddish brown 5YR \ 4 ;ini:iilar and rounded
sand with clay skins Fuoa 45 -601with rill's
46 Dark red Z-5YR 3/6 sand
27 m Reddish yellow SYR 6/6 sand with
pinkish white SYR 8/2 to pink K '3 carbonate fraebulcs
ciray 5Y 5/1 to greenish gray 6/1
9-7 m BLANCHETOWN CLAYBlacky with shiny pad laces with tiypsinu X- '
•
- — - -
—'
hemispheres and needles
Ostracotl layer
m Yellow CHOW1LLA SAND
Pvrolnsite layer / « '•.
'
0-5)3 HJ Sandy BLANCHETOWN C LAY._—
__
Lagoon level
Fig.21-—Measured section of cliff on the E. side
of Nampoo Station homestead beside
track to river level. Site 4.
point the section is about 22 m thick, but W.of the homestead the cliffs are 36- 6 m high
and the surficial reddish horizon is 6-4 m thick
where measured. This consists of reddish brownand greenish mottled clayey sand with thick
calcrete-bearing soil at the top. These sedi-
ments are interpreted as altered Blanchetown
Clay. Whether the Blanchetown Clay continues
below the river level is not known. For the pre-
sent the whole Nampoo section at Sharp Point
36 6 m thick, is referred to the Blanchetown
Clay. The dolomite members in it will be dis-
cussed later. There are intercalated bands of
sand and clayey sand. The lowest bed could be
Moorna Formation.
The top of the section near the track was
examined in more detail, and the following in-
formation obtained:
12 -7cm Reddish-brown (5 YR 5/4) angular androunded sand with fine clay skins forminga hardpan (Al horizon removed by erosion).
The hardpan tends to form verticle edges
on the cliff and is columnar. Penetrometer0-2 kg/sq cm merging rapidly to
46cm Dark red similar sand (2-5 YR 3/6), blocky
and forming a slope of 45°-60° with rills.
Penetrometer varied round 1 kg/sq cm2 -7m Reddish yellow (5 YR 6/6 but varying with
amount of carbonate and faint mottling)
sand. Numerous carbonate nodules, pinkish
white (5 YR 8/2) to pink (8/3). Boundarydistinct
9 *7m Gray (5 YR 5/1) to greenish gray (5GY 6/1)
claystone, blocky with shiny ped faces
(secondary clay). Gypsum crystals in form
of hemispheres and needles (BlanchetownClay)
Slip material covers the rest of this section,
but W. of it 60 cm of Chowilla Sand (PL 7,
fig. 2) underlies Blanchetown Clay (PI. 7,
fig. 1 ) with a thin ostracod layer at the base of
the latter. Below the Chowilla Sand is 25 mof clayey sand, the top of which has under-
gone varying degrees of silicification, inter-
preted as a function of the Karoonda surface
pedogenesis. The maximum is a hard band near-
ly a metre thick that stands out firmly on the
cliff profile. E. of the homestead the ground
surface slopes down to near river level where
a gulch cuts the cliff. Here a thick silicified
layer has been quarried. In the gulch a 12 cmband above the sandy layer is crowded with
ostracods. Determinations of fossils are given
in the section on palaeontology.
Roberts (1968) has provided information on
this formation gained by drilling in the Waikerie
district, S.A. He suggests that the Blanchetown
Clay was once over 24 m thick there, but has
since been eroded. The Bungunnia Limestone
occurs capping the Clay on two of the three
hills of clay buried under aeolian sand, but it
also occurs in one of the low areas. Thus either
the limestone is not one stratum laid horizon-
tally, or it has been faulted (which seems un-
likely).
The Blanchetown Clay is sparsely fossilifcr-
ous. At Warrakoo (Lake Victoria) Station
(Fig. 4) NE. of the homestead, the high steep
cliffs have two harder bands in this formation
near the top. They consist of siltstone and where
laminated contain in places soft plants with
simple branching, which are apparently water
plants. Ostracods are the commonest fossils,
usually crowding bands 5-15 cm thick. Gen-erally, these layers are at the base of the clay
deposit following a sharp boundary from sand,
suggesting shallow waters. However, they also
occur in Ienticles in the mass of the formation.
When examining the thick exposure of Blanche-
town Clay on the N. bank of the Murray River
at Sharp Point W. of the Nampoo Station
homestead, I saw what appeared like pieces of
tissue paper. Examination under a lens showed
that they were a layer of ostracods one shell
thick. The layer was then found in place.
34 EDMUND D. GILL
Mollusca have been found only in a brownsandy clay at the N. end of the large erosion
amphitheatre on Noola Station (Fig. 14). Onthe N. wall of the N. gulch complex thousands
of small Corbicuta occur with all sizes fromyoung to adult. The bed outcrops in three
gulches and the separating interfluves over an
area approximately 60 m by 30 in. The valves
are rarely together and are poorly oriented.
They are reminiscent of the banks of similar
shells found now on the floor of Lake Mcnindee(on the Darling River system, used as a
reservoir) when it is dry. A similar lens wasseen further S. in the same erosion amphitheatre
on Noola Station.
Fish spines, vertebrae and other fragments
were found in the same area in the Blanche-
town Clay and in sandy bands included there-
in. Oblique burrows about 3 cm diameter
were noted in the latter. At Bone Gulch onMoorna Station bones of a large marsupial
were discovered in the base of the BlanchetownClay.
At Triple Swamp (Fig. 22) on MoornaStation is a gulch with beds of dolomite to bedescribed later in this section. On the E. side
of Triple Swamp gulch are two beds of dolo-
mite in Blanchetown Clay, as shown in Fig.
23. The cliff is almost vertical and the face be-
tween the two dolomite beds slightly concave.
A pebble of milky quart/, about 18 mm by 8
mm was found in the position indicated in Fig.
23. (See also PI. 8, lig. 1). It is very well
MOORNA l \si
o\i;o\\ 1 ,\kl
Old Ron
I L'lU'C
IH)[ i>Ml If GUI ( II
A\ | !<;;>,i SWAMP'
Fig. 22—Map of localities—Moorna F. oxbow. TripleSwamp, ami Hone Gulch. Traced from air
photo Moorna Station, W. of Wcntworth,N.S.W. Site 12.
i i'-!h alive era; !.\
i
iti \\< HETOWN ( LAY
i) ( ., i'|
|
[ ( ,i | ,\] M |
ii <<p m Bl w III WAV \ ( i \\
m.Av in low s i i v.
Fig. 23— Occurrence of pebble in clay sediments at
Dolomite Gulch, Triple Sw^imp (Fig. 22),Moorna Station, W. of Wenlworth, N.S.W.Site 12.
rounded and has a slight polish. Also it is com-pletely out of character with its enclosing matrix
which is clay, the sediment of the lowest dy-
namics of all. The pebble was indubitably in
situ. The narrow end was protruding, it wasmore than half buried, and had to be dug out of
the hard clay. A fine layer of secondary minerals
lined the cavity from which it was taken. Forexample, pyrolusite was on the wall of the
cavity and in corresponding places on the peb-
ble. As the pebble is sedimentationally out of
place, it could only have reached this position
by being floated in. Pebbles of this size are al-
most unknown in this region of low declivities
and fine sediments. For this and other reasons
it is unlikely to be a drop pebble from a float-
ing stump. The most probable explanation is
that the stone is a gastrolith from a large bird
or reptile. On the interpretation of the dolomitegiven in this Memoir the water would beshallow, but even if deep the pebble could befrom a floating cadaver. Many smooth pebblesfound on clay pans at the present time appearto be explicable only as gastroliths, and are
probably from emus.
(e) Blanchetown Clay, Dolomite and Lime-stone Members
Triple Swamp (PI. 9, fig. 1) is a flat
swampland that is covered by Murray Riverfloods. The sediments are greenish gray clayeysands to sandy clays referable to the Coonam-
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 35
IS m Red Sand
046 m BLANCHETOWN CLAY0-46 in White sand and dolomite
061 m BLANCHETOWN CLAY0-91 m Dolomite and white sand
18 m Yellow CHOWILLA SAND
61 in Bunds Bl. Clav and Chowilla Sand /— —
-
0-86 m Dolomite and clay
1-8 BLANCHETOWN CLAV
1 I I I I I l_L
I>3 Brown sand and clayey sand
3-2 m Noi seen
Hillwash
Swamp level
Fig. 24—Measured stratigraphic section at DolomiteGulch (Fis. 22), Moorna Station, W. ofWentworth, N.S.W. Site 12.
bidgal Formation. This canegrass swamp dries
out and cracks deeply in summer. In places
there are sand rises with Eucalyptus largiflorens.
The cliffs behind permit one to view a long
section of the stratigraphy which consists of
Blanchetown Clay and Chowilla Sand, but with
many variations. The cliffs in the vicinity of
Triple Swamp Gulch are unusual in that three
white layers (PI. 9, fig. 2) glare in the strong
sunlight of this semi-arid region. The white
rock is dolomite, pure in some places, and
sandy in others. These white layers stand out
against the red soil at the top of the section and
the greenish gray (5 GY 6/4, 7/1) to gray
(5 Y 6/2) Blanchetown Clay. The dolomites
are dense and so stand out in the cliffs (PL 8,
fig. 1). In the gulches they form vertical faces.
The rock was too hard for the penetrometer
used, but the Blanchetown Clay gave readings
of 16-20 kg/sq cm. The red sand at the top of
the cliff gave 4-8 kg/sq cm. and the topsoil 2-
4 kg/sq cm. The topsoil is a later addition here
because an Aboriginal midden was found be-
tween the A and B horizons.
A second striking occurrence of dolomite is
found at Sharp Point on the N. bank of the
River Murray at Nampoo Station. Here again,
the beds lens out, but nevertheless reach a
maximal thickness of 12 m. Here too, there
is association in places with sand. The succes-
sion and variation is shown diagrammatically
in Fig. 25. On the other hand, there is a dif-
ference in that opal is associated with the two
dolomite layers at Nampoo. The two layers cut
out to the E. then a T2 m band appears near
the Nampoo pump, cuts out then reappears
further E. as a sandy dolomite to dolomitic
sand. Silcrete is associated with the sand and
not opal. There is a slight dip to the E. (1 in
400), and it could be that the second dolomite
band is hid from sight to the E. The changes
to the W. are shown in Fig. 25.
Messrs. Sharp and Howells Pty. Ltd. obliged
with an analysis of dolomite from the lower
layer at Sharp Point, with this result;
Multicoloured el:i>e> sandsjn
-_siiikK ila\s•
.1
Fig. 25—Sketch of successive headlands in dissected cliff at Sharp
Point (Devil's Elbow), Nampoo Station, E. of Cal Lai,
N.S.W. (Fig. 4). N. bank of Murray River. Site 4.
36 IDMUNI) D. GILL
CaCOaMi- COsSilica clc.
by difference
54 5 per cent
42-5
3
100.0%
Dr. E. R. Segnit (as reported in a paper
in I his Memoir) made X-ray and D.T.A.analyses of llie dolomites, He determined
cristabolite and tridymite. The opal beds have
caused a change in slope of the cliffs at Sharp
Point and resulted in a small headland jutting
into the river. This more resislanl rock has re-
sulted in a very sharp turn in the river here,
hence the name. Early river pilots called it the
Devil's Elbow. A section of the little headland
is given in Fig. 26.
While the whitish sand below the Chowilla
Sand is shown as horizontal, it is disturbed in
places as much as 15", apparently by the churn-
ing of the clays. Slickensides are common in
this /.one. These relationships are shown in
the field sketch made at Sharp Point repro-
duced in Fig. 27.
The opal occurs both in masses as shown in
Fig. 27, and in nodules which are generally
I'. ill nlh*; »\ (i I Mind; lt[ \\t ill I OWN < I \1
I. \Hi II >NDA I I RR \IN
' I .in I'.ik )t.ihivi S> H i ..m.i t,i -..ii.i um
I, \m n IND \ i'i i» '1)1 u\i
! m i' iti i [!nv|l
i i; « Ih »V, ii l \ s wii
r
> ,i(i-. ,i teil
In uh Wliii.- M s I.
.,,,,., iwll u.ulial .iii.l i. . .m.i .. i i —! in While i'i 'I i'Mi
1
1
plus upat with
it] \\i Ml mux i i u
9 1 W til \\< III U t\\N < I U .,,,,1 mnUu ..] ni clays in ~ "
wnd> vl.n ;ru) \\ h I ,i,ui.j;.n \ i. .1 hmwn ' n R 3 H
brownish ytlUw \\>\ K : k! scr) tlusl > rvil ' !R b K, duel red ;I
SR !I rvdvltsh bl n i
L'U
Rlvei Mum.\ knd .ii innc ol survej u.,-.,: n0 l mtii
Fig. 2<> Measured section at Sharp Point, NampooStation, E. of Cal ! al, N.S.W. N. bank ofMurray River. Site 4.
Fig. 27—Interpretation of dolomite andclay at Sharp Point, NampooStation, E. Of Cal Lai, N.S.W.Much of the dolomite hasbeen replaced by opal. Site 4.
bulbous but with some little sharp points at
irregular intervals on the surface, which is
white. The upper dolomite bed at Sharp Point
has massive rock, replaced to varying degrees
by opal, brecciated (and rcccmcnted) opalized
dolomite and nodules. The opal is a highly
vitreous translucent (clear, brownish, bluish,
pinkish) to opaque (mostly brown). Just abovethe upper opal bed is a finely laminated band(06 cm thick) rich in ostracods. Laterally, this
opalized dolomite becomes in places a siiicified
sandstone.
The opal/silica deposits are regarded as
pedogenic. The Chowilla Sand below the upperdolomite layer is mottled. The clayey sand be-
low the lower layer is also mottled and contains
rhizomorphs of quartz plus opal (E. R. Segnit
det) There is often siHcification without opalat the same stratigraphic horizon as at Cal Lai,
the Murray River cliffs at Kulcurna Home-stead, on Salt Creek and elsewhere. On Salt
Creek opal is deposited in Blanchetown Claybut in the same cliff at the same horizon only
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 37
V Red unconwjiidittcd sand
KAROONDA It-.kRAIN (rvhwnctl iwul rv-bviridd
Yelfov c IK)\\ ]|.! \ SAM)
K \KOONDA PEDQDERM
White leachud sand
BLANC IILIOWN t I \V
Fig.
Graj clayej
PARI! l,A SAND
28—Section of E. bank of Salt Creek towardsthe S. end, on Kulcurna Station, Cal Lai,
N.S.W. Site 2.
10 m away where sand has replaced the clay,
silcrete is deposited. Also sand just above the
opal has rhizomorphs, as shown in Fig. 28. In
some of the gulches at the S. end of Salt Creek,
masses of silicified sandstone have fallen to the
floor of the gulch and these provide an excellent
opportunity for studying the branching rhizo-
morphs. There are also burrows, which are not
branched, do not vary in diameter, and have
some kind of segmentation of the infill.
Although no dolomite layer was noted
in the Salt Creek section, a bore put down E.
of the creek in connection with a salinity sur-
vey penetrated a dense fine-grained rock. Bythe time I examined the site, none remained,
but Mr. Hans Hansen of Kulcurna Station told
me that it appeared the same to him as a lens
of dolomitic rock (which he kindly drew to
my attention) on the cliff behind Kulcurna
Homestead. For relation to Salt Cr. see Fig. 4.
The dolomite has suffered minor erosion in
some places, while at others it remains only as
fragments in sediments. The former is il-
lustrated by Fig. 29 which shows a section on
Nampoo Station E. of Sharp Point. Fragmental
remains can be illustrated by the W. end of
Fishermans Cliff section (Fig. 30) where pieces
of gravel in the Moorna Formation are covered
with white coatings. These worn coatings drew
attention because they looked like calcium car-
bonate and so the first appearance of the pre-
sent semi-arid climate that amasses such carbon-
ULANUII.IOWN CLA> - — — — —0.33 in i itu' greenish gray sand -'..-; - * . .
Coarser ul bastf, wllfa "hituUmI pio.es or dolomite -
«-an^-^0-3 ni 111. Clay
2 m Wink' solid DOLOMITE
Sharp boundary
: .in III AM III-TOWN CLAY
fe
Fig. 29—Section in cliff at Nampoo Station homesteadpump, E. of Cal Lai, N.S.W. ,
providing evi-
dence of erosion of dolomite. Site 4.
ate. However, Dr. Segnit identified the mineral
by X-ray as dolomite, so a gravelly dolomite
must have been formed then eroded to pro-
vide this sediment. Carbonates are so abundant
in this climate that almost any rock effervesces
on application of cold acid (as did the dolo-
mite) so this field test can be misleading. At the
E. end of Fishermans Cliff the section is as in
Fig. 31. A layer 0-6 cm thick, 0*6 m below
the top of the Moorna Formation, has small
0-9 in Peiiociil vvilli utrhomilc fracbutcs
46 in BLANCHETOWN CI \V
Pinhcnii?; of gypsum on ped lucvs /—
.
1 5 in CHow II i A SAM)Mottled iiL'jn ('my and
Reddish SYR 6/6 Ink- sand
2 1 m Coarse CMOWTI I A S\\D / -
leasts i<> lightly sillcrfied ligfti gray N/
ami reddish yellow 7-SYK 6/6
A 6 MOORNA FORM VTTON
Li-lit:
m,iv \7 8 io white sili to sand
Icwl
Fig. 30—Surveyed section W. end of Fisher-
man's Cliff, Moorna Station, W. of
Wentworth, N.S.W. N. bank of Mur-ray River. See Fig. 31 for E. end.
Site 13.
38 EDMUND D. GILL
In un pilll Mfll
°^ oi BiCulvreu; frswbulei and cartlny carhrmatc
3 in InlJUisJi hiuwn : SYR 4,-1
&LAN( HETOWJN l l \\
_ _ —3*6 ni "i.vni.li ru\ lll.W HETOWN C I W
>>>( laycj '-..in>i i L-.i.uiation CltowiUu Sjiui in hi. Clu;
Disanilor'itiilv
3 t in \\ hiic swnJ with osli'twoUs
1 tin Band with pyrolu&iic Limkrlies
I cm Band witb dolomite pebbles06 in from tap
MOORNA I GRM \T(ON
Fig. 3 1—Measured section at the E. end ofFisherman's Cliff, Moorna Station,W. of Wentworth, N.S.W. Lateraldifferentiation can be seen by com-paring this section with that fromthe W. end (Fig. 30). Site 13.
rounded to angular pebbles of dolomite. Justbelow this band is one rich in pyrolusite about8 cm thick. The same detrital dolomite bandassociated with pyrolusite occurs to the NE. in
the gulch at the NW. corner of Moorna Oxbowlake. So the dolomite was once more widespreadthan appears by present outcrops. In a rapidreconnaissance of the Murray River upstreamfrom the study area, dolomite (E. R. Segnitdet.) was found on Tammit Station about14-5 km from Euston. In the Sandhills Paddockabout 16 km WSW. of the homestead, Mr. O.Grayson showed me the section given in Fig. 32,where the right bank of the Murray River re-
veals:
0-3m Gray loam5 -5m Very pale brown 10 YR 7/3, with yellow
mottles 2 5 YR 7/6 clayey sand. At base,rluzomorphs, carbonate patches, and dolomitenodules forming a berm in river bank
6m Micaceous clayey silt, white 10 YR 8/1 merg-ing into mottles of strong brown 7-5 YR5/6-8
1 -f- m Clayey sand with iron oxide mottles, becomingironstone in places. Forms reef at Tammithomestead.
Similar sections were seen in other places onthe river in this area.
Firman (1966, 1971a) has described theBungunnia Limestone in S. Australia as a
U 3 Oruj 1<m;ii
5 .(> in Vary pQk brown loYk 7/3 i
.-!.
s.hkI witli niuitl.-, of yellow 2-5Y 7/6
\.„M, liuicj piitchci .in,!i .n bnnflli
ihi/ninorpli
n Micaceous d*ro i ill white I0YR'
merging )ntu muttlcs oi strong brown
Not seenR 5 6-8
32—Measured section on N. bankof the River Murray, TammitStation, 16 km from Euston.N.S.W., in the Sandhills Pad-dock. The carbonate is dolo-mite.
formation succeeding the Blanchetown Clay andoccurring E. of the Morgan fault line scarps,
also W., N. and NE. of the Pinnaroo Block. It
is a "micrite flaggy or banded, oomicrite, oolitic
algal biolithite, calcilutite; generally dolomiticand containing ostracodes". This formationdoes not occur in the study area, but there arelenses of similar lithology in the BlanchetownClay. They are related to the Nampoo andTriple Swamp dolomite members but are litho-
logically distinct. The members described aboveare pure white dolomite, extremely fine in tex-ture (Segnit, et al. this Memoir)), and nofossils have yet been found in them. By contrastthe Bone Gulch dolomite member and theMorkalla dolomitic limestone member are light
brown, much harder rocks (the latter is usedfor paving paths) and highly fossiliferous, beingpacked with ostracods and other evidence ofbiologic activity.
On the W. side of Bone Gulch on MoornaStation, N.S.W., erosion has stripped Blanche-town Clay from the top of the dolomite which inturn has protected mottled Blanchetown Claybelow it. The rock is pale brown, whitish orpinkish in places, and of low specific gravity,due to the large numbers of ostracods (Ilyocy-pris, Candona and Diacypris). The bed is only1-3 cm to 5-4 cm thick, but extends 140 malong the river, and intermittently for another60 m. As the deposit has been eroded by theMurray River, its original extent is not known.The deposit indicates shallow lacustrine con-ditions. It is sandy in places. The ostracods are
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 39
dolomitized, so secondary dolomitization is
involved. The matrix is olive gray 5 Y 6/2 and
the mottles reddish-brown 2 5 YR 5/4 andwhite 75 YR N8. The white is due to gypsumcrystals and carbonate. Rosettes of gypsum upto 12,5 cm occur but most of this mineral is
in small pieces 0-5-2-5 cm diameter. The clay
becomes greener with depth. Professor R. H.Tedford described (inlitt.) "ostracod coquinas"
in the water supply canal cut between LakePamamaroo and Lake Menindee to the N., and
others are reported elsewhere in this paper.
In the Morkalla district of NW. Victoria are
the properties of Messrs. B. Nunn, R. and L.
Gray, and F. Heitmann which have on themflaggy dolomitic ostracodal limestone. Wherethick enough, and near enough to the surface,
it is ripped up with a tractor and sold for mak-ing ornamental paths. Dams on the Nunn pro-
perty show clay which is probably BlanchetownClay. On the Old Station dam on Grays' pro-
perty, limestone is reported at 2*5 m with clay
above. About 15 km WSW. of L. Grays'
house (PL 8, fig. 3) just E. of a N-S. fence and
S. of an E-W. fence, a scour showed ostracodal
limestone in situ (PL 8, fig. 4). At this site it
is horizontal and in three layers over a marly
clay. The limestone is a light greenish colour
inside and off-white on the outside. A section
was measured as follows:
5cm Red sand and small pieces of calcrete (hill-
wash)2cm Layer 1 of limestone, crazed on top
l-9cm Layer 2lcm Layer 3
40 -f- cm Mottled green, white and reddish brownclay to clayey marl
The limestone was followed round the base
of a red sand ridge at the same contour for
1-5-2 km, indicating that the deposit is horizon-
tal and probably extensive.
( e ) Chowilla Sand
The Blanchetown Clay (PL 7) and the
Chowilla Sand (PL 6, fig. 1) are the two most
characteristic formations of the study area, but
they contrast strongly. The former is pre-
dominantly clay, while the latter is pre-
dominantly quartz, usually well-washed. Theformer is a low dynamics lacustrine sediment,
while the latter is a riverine channel deposit.
Similar deposits occur in the Holocene Coonam-bidgal Formation where the loose channel sands
provide the source for the structuring of chan-
nel border dunes. There are parts of the river
at present where the dynamics are such as to
wash clay from sand and provide sufficient of
the latter to build limited beaches and dunes,
e.g. in the Mildura area. The BlanchetownClay and the Chowilla sand interdigitate to a
remarkable degree in this area, and (as is to beexpected) sometimes grade into one another.
Firman (1971a, p. 57) provides the origin of
the name of this formation and its type locality.
He gives the distribution as from Merbein in
Victoria to Berri in S. Australia. In a rapid
reconnaissance of the Murray valley upstream,
Chowilla Sand was noted at Monak in N.S.W.in sand quarries on the right bank of the
Murray River and at Robinvale on the left
bank, where the formation is partly silicified.
The lateral differentiation of the Chowilla
Sand is illustrated by Fig. 33. At the windmill
T Monrna Sian'un windmill Corner in cliffs
RciJ sand with taliri-lc I'racbuics,*
rReti sandj? - —T'-tSOIL
CHOWILLA SAM)
SITE 1 1-
, CHOWILLA SAND" BL. CLAY
Fig. 33—Section of cliff N. of Moorna Station homestead, W. of Went-worth, N.S.W. Sketched from measured sections and cliff traverse
to show lateral differentiation. Site 11. cf. Fig. 34.
40 EDMUND D. GILL
end of the section are some bands of silcrcte
which the Aboriginals could have used for im-
plements. At Moorna Station homestead exten-
sive silcretc outcropped at river level before the
building of locks raised the level of the water
there, Mr. A. T. Honncr informed me. Thestone was used in the building of the homestead.
At the corner marked on Fig. 33, N. of the
homestead, the section was measured as fol-
lows:
5m Red sand1 Oni Same with calcrete nodules, white 5 YR 8/1
to pinkish white 8/25 '3m Greenish olive gray 5 Y 5/2 Bhinchclown Clay5 4m Yellow C howilla Sand to river level.
The Blanchc'own Clay has shiny ped faces
with line selenite crystals, and oxidizes to light
gray 5Y 7/1-6/1. In the middle is a narrow
band of white 5Y 8/2 sand with some clay
laminations. The lop 5-10 cm is dark with
pyrolusite, which is common as a secondary
mineral in the sand formations. In the basal
Chowilla Sand more pyrolusite occurs, and a
piece of fossil bone was found. The bed should
be prospected for vertebrate fossils. This is the
same horizon as that in which occur the plenti-
ful fossils at Bone Gulch.
Figure 34 provides a diagram of part of the
Triple Swamp cliffs. It relates the important
Bone Gulch (with nearby ostraeod dolomite)
to the Triple Swamp Gulch with its three dolo-
mite bands. It also shows the extensive in-
lerpenetration of ("howilla Sand and Blanche-
town Clay.
The two illustrations given of inlerpenetra-
lion, and the many sections with Chowilla Sand
provided earlier in the paper, prove that this
formation cannot be used as a time indicator
but only as a fades indicator. In such conditions
of lateral change it is difficult to correlate one
section with another because of lack of datum
planes. However, these are provided in the
study area by pedoderms and fossils. Both of
these require further study and refinement, but
this will be discussed in the section on chron-
ology. Another factor that can cause difficulty
is the presence of disconformities and inter-
formational erosion. The pedoderm is a dis-
eonformity in that a significant amount of time
was involved in its formation. It represents a
period of non-deposition. There are evidences
of interformational erosion (e.g. Fig. 29) but
these are on a minor scale, being usually small
channels. They are therefore not very significant
with respect to the stratigraphy. In the Paringi
(N.S.W.) section two such channels occur in
the top of the latcrite. A minor one occurs in the
top of the Chowilla Sand at Bone Gulch.
Although the Chowilla Sand dominates the
section in some places (e.g. 33) it is absent or
very reduced in others. Fig. 35 shows a section
in a gulch running N. at the N. extremity of
the Salt Creek cliffs in the Scrub Paddock of
Kulcurna Station (originally called Hypurna,the Aboriginal name for the creek — J. Hig-
gins).
(f) Moorna Formation
This is a new formation hereby proposed for
an assortment of mostly unoxidized riverine
deposits, gravels to silts, that occur below the
Blanchetown Clay in the very broad andshallow Lake Victoria syncline. Because the
Parilla Sand in this area is warped below river
level, the only other formation occurring be-low the Blanchetown Clay is Chowilla Sand.This well-washed oxidized well-sorted channeldeposit is readily distinguished. Unlike the
Fig. 34- Sketch of the cliff section between Bone Gulch andDolomite (iulch (surveyed sections) to show interdiuita-
lion (if Blanchetown Clay and Chowilla Sand. c. Fig. 33.
For localities see Figs. 1, 4, 22 Site. 12.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 41
l. ^consolidated reel aund
KAROONDA TERRAIN0-6 m Siliciiial C'HuWil I A SAND .
KAROONDA PEEOD1 KM
!2 m BLANCHETOWN ( I \Y/"—
'
Hard b ind
1-5 in PAR.H I. A SAN!)-'-'
Fig. 35—Measured section of gulch in ScrubPaddock, Kulcurna Station, CalLai, N.S.W. Site 2.
Moorna Formation, it rises through the
Blanchetown Clay, divides it, and caps it. TheMoorna Formation always maintains its
stratigraphic position.
The type section (Fig. 30) is on MoornaStation (from which it derives its name) at the
W. end of Fishermans Cliff (Fig. 4, Site 13)so called by us because a fisherman has his hut
at the E. end of the cliff. The paddock above
is called The Selection by the Moorna manage-ment.
At the W. end of Fishermans Cliff (Fig. 30)the Moorna Formation consists of poorly-
sorted gravel to coarse sand with the fragmen-
tary bones of fish and marsupials, but at the E.
end of the cliff (Fig. 31) it has laterally graded
into laminated clayey silts and other fine de-
posits. The former is interpreted as a channel
deposit, and the latter as a floodplain deposit.
At the W. end the Chowilla Sand (with a fossil
tortoise) is interposed between the MoornaFormation and the Blanchetown Clay. TheChowilla Sand is not quite typical because
a lateral differentiation is beginning that be-
comes clear at the E. end of the cliff. For
example, besides the usual well-sorted light
brown sand there is some light gray (N7) with
large reddish yellow (7 5 YR 6/6) mottles,
and some is coarser than usual. Below is 46 m(and perhaps more) of Moorna Formation con-
sisting of light gray to white (dolomitic)
N7 to N8 coarse sand to gravel. It is clayey in
some places and clean in others. A bore is des-
irable to determine the thickness of the deposit,
and to discover if Parilla Sand lies below it. An
undisturbed core through to the BookpurnongFormation would be enlightening.
At the E. end of Fishermans Cliff (Fig. 31)the Chowilla Sand thins to 9 m, is clayey, andis greenish gray where unoxidized, i.e. it has
graded into a sandy phase of the Blanchetown
Clay, at least 7 m of which overlies it. A sharp
break at the base appears to be a disconformity,
and may be the Karoonda Terrain. The under-
lying Moorna Formation consists of con-
solidated sediments that form vertical cliffs at
the section site. The top is 79 m from the top
of the cliff.
The succession is:
Top 0-6m Very pale brown sand 10 YR 7/3-4,
some very hard (20 kg/sq cm) andsome very soft (2 kg/sq cm)
6cm White sand and pebbles of dolomite9cm Quartz sand with black pyrolusite
The above two narrow bands have alocal dip of 40° for 0-6m, which withthe detrital dolomite is evidence oferosion. The eroded band is 28-60cmof white to very pale brown (10 YR8/1-4) unstratified sand cemented in
places with silcrete
3-4m white sand to clayey sand; 76cm ofthis layer is finely laminated, 84laminae being counted in 10 cm.Penetrometer measured 12-20 kg/sqcm
2-7 -f-m Intercalated very pale brown sand(10 YR 7/4) and reddish brown(2-5 YR 5/4).
The Moorna Formation here is thus 163+ mthick. The thickest outcrop noted was at
Merbein, where the left (W.) bank of the River
Murray at the scenic lookout between the
Winery and the Pumping Station reveals the
section shown in Fig. 36. Here below mottled
Blanchetown Clay up to 3 6 m thick is 4 3 mof Chowilla Sand. On the section line there is a
conglomerate of masses of silcrete, quartz
gravel and yellow sand lying on an uneven floor.
Further S. towards the Pumping Station, a
silcrete layer is in situ, and obviously this layer
has provided the conglomerate. It is the
Karoonda Terrain, and the silcrete is a
pedoderm.
The Moorna Formation consists of over
17 9 m of off-white to light gray, very poorly-
sorted, current-bedded sands to gravels with
varying clay content, sparsely micaceous, pre-
42 EDMUND D. GILL
Middens
3 4 m BLANCHETOWN CLAY mottled
3 7 CHOWILLA SAND
KAROONDA TERRAIN0-6 Silieified sandstone
KAROONDAH PEDODERM
18 m MOORNA FORMATION - *__l : :_L-i-r :
White sand to fine gravel with current
bedding
1 m ironstone
Murray Rrvei Level
Fig. 36—Surveyed section of the cliff on theleft bank (W.) of the Murray Riverat The Lookout, Merbein, Victoria.
dominantly clear quartz but some milky, withrounded to angular grains, cemented in places,
rhizomorphs present in the upper 3 m, un-oxidized. By contrast, at the base (at river
level) is 0-6-1 -8 m of soft ironstone throughwhich water seeps to the river. The ferruginized
sand is included in the Moorna Sand. Whatfurther thickness lies below is unknown. A boreto determine this and discover what formationlies below would be useful.
Referable to this formation are the strata
below the Blanchetown Clay in the E. Moornaoxbow lake section (Fig. 6), of which the detail
is as follows:
Top 51cm Firm (20 kg/sq cm) red-brown (2-5 YR4/4 and some lighter shades of clayeysand with pyrolusite
0-6cm Light brown (7-5 Y 6/4) do. withdetrital dolomite
18cm As at top but more light brown in colour0-6cm Dolomite20cm As 18cm layer above13cm Stiff clayey sand. Same colour as above
but with purplish bands. Gypsumrosettes 1 . 5-3cm diameter.
Thus the basic regional stratigraphy consists
of Bookpurnong Formation (oldest) LoxtonSands, Parilla Sand, Blanchetown Clay,
Chowilla Sand and Moorna Formation. Formost of the study area only the last three out-
crop. At this stage of geologic development,
the Lake Bungunnia system became defunct
as the Murray River cut down into this succes-
sion of strata, and drained them to base level.
The formations listed below as further re-
gional deposits are thin surficial ones overlying
the foregoing basal units. This is why the
Blanchetown Clay forms the surface of much of
the country and is widely used for excavationof tanks (water reservoirs).
(g) Bakara Pedoderm (Firman 1963, 1966,
1971a).
This is discussed in the accompanying paperon palaeopedology.
(h) Woorinen Formation (Lawrence 1966,Finnan 1971a).
This comprises the E-W dune system des-cribed earlier (Figs. 9-10).
(i) Lovedav Pedoderm (Firman 1966,1967b).
This soil is developed on the WoorinenFormation dunes, and is discussed in the ac-
companying paper on palaeopedology.
(j) Bunyip Sand (Firman 1966, 1967b,1971a).
This formation occurs as dunes, or as aveneer of red sand on the plateau in which theriver tract is cut. About 08 km S. of TalgarryStation homestead is an example of a red sanddune belonging to this formation. It was erodedby winds during the 1967-8 drought so that its
structure could be examined. The sectionsexamined showed a homogeneous compactedred-brown sand. Aboriginal middens were as-sociated with the dune, both on it and in it.
The internal middens could be used to date thedunes.
On the hillslope above the Salt Creek mea-sured section (Fig. 20) Bunyip Sand forms adeposit that is partly hillwash and partly aeolian.A section in a gulch showed:
Top 13cm Loose red sand10cm Weak red (10 YR 5/4) sand forming a
hardpan held by clay skins36cm Deeper red (10 R 4/3) sand
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 43
60cm Weak red sand lightened in colour bypatches of earthy carbonate and nodules;also burrow fillings and in a few placesthin seams down joint planes
60cm Mottled weak red (10 YR 4/6) and red(2-5 YR 5/6) and yellowish red (5 YR5/6) with off-white patches of carbonateand white specks of gypsum. Latter mostlyin lower half of this horizon. Also roughsurface non-magnetic ironstone concre-tions up to 1-3 cm diameter
(k) Yamba Formation (Firman 1971a, Law-rence 1966).
Gypsiferous clays and sands in playas and
low dunes. The gypsum is commonly of the
"seed" and "flour" types. An example in the
study area is the Morkalla district. Lawrence
(1966) refers to a number of such areas.
3. River Tract Formations
The Murray River is incised deeply into the
regional formations so that the present flood-
plain is about 37 m below the general terrain
(Fig. 12). In the section on geomorphology,
the tract in the area of study has been described
as a pattern of higher red country and lower
gray country (Fig. 5). These two types of
country are surface expressions of two forma-
tions which will now be described, along with
another formation proved by the Chowilla Damsite investigation, viz.:
Top (c) Coonambidgal Formation (Gray
country)
(b) Monoman Formation (underlies c)
(a) Rufus Formation (Red country).
In addition to these riverine formations,
there is the aeolian Lake Victoria Sand.
(a) Rufus Formation
This is a new formation here proposed and
the type locality is the outlier shown on Fig. 5,
where vertebrate fossils have been found that
prove a Pleistocene age. The site is between the
Wentworth-Renmark Road (at a bend in it)
E. of Lake Victoria and N. of Frenchmans
Creek on Dunedin Park Station, N.S.W. The
name is after the nearby Rufus River which runs
from Lake Victoria to the River Murray. The
lithology is a clayey fine sand, and the fades
riverine, as is shown by the nature of the sedi-
ments along with the flatness of the tops of the
outliers. The latter were obviously once a con-
tinuous river terrace that was corraded and dis-
sected. The outliers themselves have suffered
some dissection—a mark of antiquity in this
low rainfall area. The red terrace residuals
stand up to about 7 m above the surrounding
gray Coonambidgal sediments, and aprons of
hillwash are common. The sandy fraction maybe re-worked into local small dunes. The vege-
tation on this formation is Eucalyptus largi-
florens, Atriplex, and such shrubs, then an
under storey of herbs. The flora is renewed by
floods. With it is associated a fauna of birds
(including emu), marsupials, reptiles and in-
vertebrates. Aboriginal middens, ovenstones
and skeletons are found on the Rufus Formation
residuals but none was found in the surround-
ing Coonambidgal floodplain. Three Velesunio
middens were noted at the S. end of outlier 1.
Middens also occur on the reworked red sands
at the borders of some outliers. Small rounded
polished pebbles occur, and these are thought
to be emu gastroliths. The stratigraphy wastested by a surveyed section with auger holes.
Aboriginal Skeleton Situs Dunedin Parfc StationRed Pedoeal.-T7w
Auger Hole 1 ^^fl!^ ,0>v
n Mint )rmivrtTy- : -:-'. — '-^^>sJ'^"i^ ^ rioieu
SITE 10'Carbnnaie ZonTr
~""~'=^'-"——*-
~!""^-',rr~—-^^- ,
-11
"^j-3^S-Er,
."tD* zlt-S~5 _
\<\ I \ S 1 ORMA I'M *^ -i-i, •
t~ .-_—— • -»Claye> sand- ." j~~ : ".— .
-" —
-
|Scales
iiL_: i ; _i • •; : : i_^ L_ -—
I'
Pig, 37—Section in Rufus Formation residual marked1 in Fig. 5. The site of Auger hole 2 wascovered by the 1956 flood. It is at approxi-mately the same level as Auger hole 1
(extreme left of figure) in Brown's Pad-dock. Keera Station, on the other side ofthe river (Fig. 38). Broken lines connectthe limits of the carbonate zone proved inthe two auger holes, but it is likely thatthe pedocal follows the present land sur-face. Site 10.
On the W. side of outlier 1, the section shown
in Fig. 37 was determined, and the bore logs
were:
Auger 1 72m E. of N-S. fence
0-20 m Red 2-5 YR 4/6 fine micaceous sand withvery little clay (about 30cm of red sandhas been eroded off this surface) grading to
44 EDMUND D. GILL
33 m Red 2 5 YR 5/8 do. with masses of hardcarbonate, grading to
41 m Very pale brown 10 YR 7/3 do. Gradualchange to
86 m Light yellowish brown 2-5 Y 6/4 almostgreenish fine sand. Sudden break to
08 m Light gray 5Y 7/2 fine sand. Decreasingamount of carbonate
0-28 m Yellowish brown 10 YR 5/8 fine sandwithout carbonate. Fairly sharp change to
58 m Light gray 5 Y 7/1 do. with yellowish
brown mottles at the top0-33+ m Same but darkened by pyrolusite. More
clayey and some mottles. Plentiful micaat bottom.
3-07 m
Auger 2 On top of residual on clay pan where about0-3m red sand has been stripped.
76m Red 2,5 VR 5/6 fine slightly clayey sand.Fine carbonate at surface and hard lumpsfrom 25cm
0-36m Reddish yellow 5 YR 6/6 micaceous fine
sand with little clay. Carbonate fracbulesup to 4cm diameter, merging to
63m Lighter in colour and less carbonate. Verypale brown 10 YR 7/3 fine sand withfraebules to 2cm diameter but usually less.
Marked change to2 08m Brownish-yellow 10 YR 6/6 fine micaceous
sand with little clay or carbonate. Lighterand darker bands present.
Auger 3 On slope at R.L. — 7 -32m.0-43m Red 10 R 4/6 moist fine sand merging to
yellowish red 5 YR 5/6 merging to018m Reddish yellow 7.5 YR 6/60-76m Pale brown 10 YR 6/3 fine sand with white
masses of carbonate. At 0-81m from sur-face small translucent crystals of gypsum(selenite). At l-22m from surface frae-
bules up to l-3cm in diameter. Disappearedby l-45m. Gradual change to
1- 27m Light brownish gray 10 YR 6/2 clayeysand with small hard pieces of carbonate.Reduced to earthy patches at l-5m anddisappears at l-8m. Merges z
79m Same with light gray mottles. Sandier at2 -85m
0-84m Light gray (2-5 Y 7/2 to 10 YR 7/2)slightly clayey sand with faint mottles oflight yellowish brown 10 YR 6/4. At —2 -6m light bluish-gray mottles dominateand a layer of pyrolusite occurs. Rest ofcore to 4 -27m light bluish-gray with 10 YR6/4 and 7-5 YR 5/6 mottles. Water at4 -27111.
Auger 4 On floodplain of Coonambidgal sediments.015m Strong brown 7-5 YR 5/6 slightly clayey
fine sand0*91m Brown to yellowish brown 10 YR 5/3 to
5/4 clayey micaceous fine sand. Gradualchange to
0-41m Mottled pale brown 10 YR 6/3, brownishyellow 6/6, and light gray 7/1 micaceousfine sand. Marked change to
0-4 lm Mottled gray with decrease to disappear-
ance of pale brown and large developmentof strong brown 7-5 YR 5/6 to dark brown4/4. Increasing amount fo pyrolusite form-imr black patches. Sudden change to
0-20m Pale brown 10 YR 6/3 do. Water (salty but
not strongly so) at 2 -08m from surface
0-36m Clayey medium to coarse sand mottledpale brown and gray colours mostly 5 Y6/1 and 5 GY 6/1. White carbonate form-ing hard lenticles. At 2 -29m from surface
oxidation colours gone0-28m Pale brown sand for 15cm then layers of
greenish gray colours 5 G 5/1 to 4/1.Some strong brown mottles at 2 -7m. Watermoderately salty.
In summary, the Rufus Formation consists
here of fine micaceous sediments fully oxidized
(red) at the surface, where by aeolian action
sand has been sifted from clay. Only one thin
lenticle (Auger 4) of coarser sediments wasmet. Below the surface brown colours dominate,
below which the sediments are in a state of
chemical reduction. Pedogenic carbonate fol-
lows the land surface so the erosion of this
bank anyway took place in the Pleistocene.
Perhaps heavier rainfall occurred then, and the
geomorphology has remained much the same in
succeeding drier times. It has been noted else-
where, e.g. Brown's Paddock, Keera Station,
Victoria (Fig. 38), that the carbonate soil onthe Rufus Formation follows the present land-
form. The soil on the Rufus Formation consists
of:
Al 30m Red micaceous fine sandA2 0-20m Hardpan of red sand plus clay but with-
out carbonateB 1-8 m Brown sand with carbonate and some
mottlingC 6-f m Mottled to gray sand.
The Coonambidgal sediments (represented
by Bore 4 above) have no developed soil profile
or accumulation of carbonate.
One outlier (Fig. 5) is horseshoe shaped,but is not a channel border dune. Its perimeterdefines a Holocene meander, and indeed suchpermit the earlier vagaries of the river course to
be worked out. The centre is not at floodplain
level, but consists of a gradual slope.
In the NW. corner of Moorna Station, N.S.W.(air photo Lindsay Run 1, No. 5112) is outlier
7 (Fig. 5) which overlaps slightly into DunedinPark Station. It is very irregular in shape andlarge, with a flat top. In Brown's Paddock on
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 45
uieid poo|j juasaJdxoq^oeta
L)snq enig
MMl'i I'M
1ji m
r ii
fit 1 li II
o>-z.
o
CO
<DCLUUJ
2 isau jus |isso-j
U0J6]3>|S |BuiBuoqv
I jsau 1UB iissoj
jUB|d oooBqox
ijsnq enja
SjOquMs UOIIB1369A)
xoq >pB|3
3*
n) ON
pues ^GjBgtlauuEijo U|
9|0M J36ny
Keera Station, Victoria, extensions of the Rufus
Formation terrace form two dune-like ridges
as shown in Fig. 38. The outcrops in this area
show that the Pleistocene terrace represented
by this formation was very wide, like the pre-
sent one. On present evidence, the overall
geometry of the river tract is not significantly
different from what it was in the Middle Pleisto-
cene (this term is purposely vague). The soil
has mobilized calcite and selenite during its
formation.
An outlier of Rufus Formation also forms
an anchor for the S. end of the Lake Victoria
lunette (Fig. 5). The sharp rather straight
line defined by the S. limit of the lunette is a
function of sand being blown past this outlier.
Finally, one of the most significant relics of
this terrace is within the Lake Victoria basin,
and between the lunette and the east wall of the
basin. This has yielded a ramus of Sarcophilus,
the Tasmanian Devil. The presence of Rufus
Formation within the basin proves that the basin
was cut before this formation was emplaced.
Its presence so close to the wall of the basin
suggests that the basin, like the rest of the river
tract, has much the same morphology as it had
in the Middle Pleistocene. These matters need
more adequate investigation.
(b) Monoman Formation
The Murray River in the study area has
scoured its tract to some depth, then infilled it
again—a mark of changed conditions. At the
Dam site (Fig. 4), over 390 river miles from
the mouth, the Loxton Sands (Fig. 18) are
incised to about 1*5 m below present sealevel,
while at Swan Reach the down-cutting reaches
about 15 m below sealevel (Firman 1971b, p.
3). The valley fill consists of two formations,
the Monoman Formation being the lower and
the Coonambidgal Formation the upper. Theformer was named by Firman (1971a and
references) after Monoman Creek in the Damarea, and the D line of bores there is the type
section (Fig. 39).
As there was danger of dam underflow
through the coarse sediments of the MonomanFormation, a bitumenous grout curtain was de-
signed to cut off water movement. A deep trial
trench was dug for experimental purposes on
46 EDMUND D. GIU.
\\ / //COON wmnx . \l I MK\I \| l< r\
V (I I II M SI H \1 I
Fig. 39—Map of proposed Chowilla Dam area to show the extreme flatness of the present flood-plain, and
the origin of the fossil finds (Fig. 40). Based on S.A. Engineering and Water Supply Chowilla Dammap, whose datum is 106-88 ft. below mean sealevel, Adelaide. The contours are in feet. Thusalthough the Murray River at the downstream end of this map is c. 390 miles (628 km) from the
sea, it is only of the order of 53 ft. (16 m) above sealevel. Site 1.
the lloodplain enclosed by the loop of the
Murray River proper at Big Bend near the inter-
section of E.W.S. co-ordinates 99, 200 N and
60,050 E (approximately Lat. 140°25S, Long.
33° 59'S). The contractors placed samples of
about 1 m :t of each of the sedimentary layers
in a row, so that those involved with the setting
of the grout curtain could devise suitable pro-
cedures. I examined these samples, and the
depths given are approximations provided by
the engineer, Mr. J. Kcndricks:
-3 m Gray sandy clay of the present fleodplain
(unoxidized)3 -4 5m Transitional, the percentage of clay re-
ducing with depth, the coarseness of sedi-
ments increasing, and the oxidation status
increasing
4 5-6 m Yellow, well washed, poorly sorted coarsesand to gravel (aquifer)
6 -7 5m Transitional, from oxidized to unoxidized7 5-15 m Chemically reduced sediments, pH c.6.5.
Mr. Kcndricks also advised that there is an
horizon of logs and stumps across the valley at
a level between 82 and 10*1 m below the
surface. From the sample sediments 1 collected
a small log and pieces of wood. Mr. H. D. Ingle
of C.S.I.R.O. Wood Structure Section deter-
mined these as:
Eucalyptus largiflorens (log)
Eucalyptus camaldulensis?
Casuarina luehmannii?
These species are characteristic of the area
at the present time. A piece of the log was dated
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 47
by radiocarbon 7,200 ± 140 y B.P. (GaK-2513), while another piece of wood from the
same fossil tree horizon dated 4,040 ± 100 y(Firman 1967b, 1971b). The latter is reported
to come from "25 ft." below the surface, a
rounded figure that probably means 7-8 m.
However, it is clearly from the upper part of
the wood horizon, so its younger age is to be ex-
pected. As these samples were wood, and ade-
quate in size, their dates may be expected to be
reliable. If so, and neglecting biologic age, a
difference exists of 3,160 years, a considerable
gap in Holocene chronology, and one associated
round the world with a small climatic change.
A disconformity is indicated, probably followed
by rapid burial of the trees to preserve them.
It is significant that in the D line of bores (Fig.
18) and the test trench, fossil wood was found
only in this zone, and plentifully. This horizon
may thus be taken for the present as the boun-
dary between the Monoman Formation and the
Coonambidgal Formation (Fig. 40). However,
more precise data are still required.
The small fossil log grew fine, acicular, white
to very pale green crystals of melanterite (de-
termined by Dr. A. W. Beasley). They are
Urn
10
15
W
?•I
SITE 1
Gray sjndy clay (unoxidi/al)
COONAMHIIKiAl, FORMATION
Ydlmvish course sand to grave! (oxidized)
Former land surface
Disconluimity with C')4 dates 4040 &7200 yr I-orestcd J loodplain
pH c.6-5
coarse sand to gravel (unoxidized)
MONOMAN FORMATION
Femur of Phascohmts gigas 16 8 m
Vertebrae of Macwpusjenagus 22-3 m
Fig. 40—Coonambidgal and Monoman Forma-
tions as revealed by the Test Shaft
(Fig. 39) and cutoff trench, Chowilla
Dam site, S.A. Site 1.
probably derived from the decomposition of
marcasite. The environment was acid, as it to
be expected where plant matter is decomposing,
so the form of iron sulphide would be marcasite
(Edwards and Baker 1951). After the scrap-
ings were taken for identification, further cry-
stals, grew on the same site, indicating continu-
ing oxidation. Melanterite often occurs as an
efflorescence on the walls and timbers of mines
in the oxidized zone of pyritic ore bodies,
especially in arid regions. Jarosite, also a pro-
duct of the weathering of marcasite, was found
infilling cavities, including cell spaces.
A black (chemically reduced) femur of
Phascolonus cf. gigas (Marshall, this Memoir),
the giant wombat, was found in the MonomanFormation at 168 m from the surface in the
test trench, and a vertebra of Macropus cf.
ferragus at 22 3 m in Bore Hole 20 on line D(Fig. 39). These fossils indicate a Pleistocene
age, and thus support the idea of a disconform-
ity at the fossil tree horizon. On the chronology
suggested here, the Coonambidgal formation is
Upper Holocene and the Monoman Formation
Lower Holocene/Uppermost Pleistocene, with
a Middle Holocene interval between. TheMonoman Formation thus correlates at least
in part with the Nulla Nulla Sand Member of
the Lake Victoria Sand that constitutes the
lunette on the E. side of Lake Victoria. TheMonoman Formation is younger than the Rufus
Formation, with a time interval between themduring which that alluvium was dissected and
deep channels corraded. This was no small in-
terval, as a very large area (Fig. 4) was eroded
and the incision was deep. As no suitable boring
equipment was available, the distribution up-
stream of the Monoman Formation could not
be determined.
The question remains as to why the MurrayRiver corraded its course, then reversed this
process and deposited the Monoman Forma-tion. Corrasion is a function of rejuvenation. In
spite of the Pinnaroo Block being a zone of
tectonic uplift, the river there is cut below sea-
level. This prompts a glacioeustatic interpreta-
tion, viz. that lowering of sealevel caused the
corrasion. If the age given in this section to the
Monoman Formation be correct, then the Last
Glacial was the time when corrasion took
48 EDMUND D. GILL
place. However, it is not so easy to explain the
infill. Rise of sealevel during the Flandrian
Transgression could account for a new base-
level but not infill to a considerable height above
it. The river bank at the dam site is about 16mabove sealevel. Because S. Australia includes
the Lake Eyre basin which is depressed below
sealevel a datum of -100 ft from mean sealevel,
Adelaide, was accepted, which has now been
corrected to 106*88 ft. This figure has to be
subtracted from the contours on Fig. 39. Thealluvium rises to approximately 200 ft. at
the dam site, which is about 28 m above pre-
sent sealevel. Moreover, the Monoman Forma-tion is a high energy deposit of well-washed
sand and gravel, while the present floodplain
consists of low energy clayey fine sand. Somefactor other than eustasy is involved, and is
probably a climatic one.
(c) Coonambidgal Formation
Both the Monoman and CoonambidgalFormations are variable riverine deposits best
accommodated under the designation "Forma-tion" rather than a rock term. Nevertheless,
in the study area the latter has a sediment range
offset to the fine side compared with the former.
The Coonambidgal Formation overlies the
Monoman Formation. The type section is the
bank of Coonambidgal Creek, at the NE. corner
of lot 75, Parish of Deniliquin North, N.S.W.(Butier 1958, Lawrence 1966, p. 548). This
Formation has time equivalents (in part at
least) on the general terrain above the river
tract in the red Bunyip Sand (or redistributed
such), the white Molineaux Sand, and the
gypseous Yamba Formation. In the Lake Vic-
toria basin its time equivalents are in the
Talgarry Sand, the very recent Dunedin ParkSand, and the colluvial fans on the scarp form-ing the W. bank. However, all writers do not
use this formational name in the same sense.
Closer definition and dating is needed.
Before the Coonambidgal Formation wasformally defined, Butler (1958) used the termto describe a "system" of riverine sediments.
His section was adopted as the type one. Therethe sediments overlie the Mayrung Surface. AtMoorna East Oxbow, Coonambidgal sediments
overlie an old land surface (Fig. 7).
While the Coonambidgal sediments are
flooded by the present river, they are not com-
pletely covered, showing that they are not com-
pletely in phase with the present river regime
(cf. Dury 1960, 1965). Pels (1964b, p. 115)
describes how the Murray follows the Coonam-bidgal sediments as far as Bullatale Creek off-
take (W. of Tocumwal), "from whence it nowtravels parallel to them and a few miles S." Pels
gives other examples of the partially ancestral
character of the Coonambidgal Formation. Its
age is Holocene. However, in the study area,
all post-Blanchetown riverine sediments are
confined to the present incised river tract. Withthe breakdown of the Pinnaroo Block barrage
(or Padthaway Ridge) the Murray Gorge wasincised. Upstream, in the slightly downflexed
Lake Victoria region, the river swept to and
fro, cutting the wide river valley that wouldmake a Chowilla Dam inundation area so large
(Fig. 4). This time of downcutting corresponds
to some of the ancient stream deposition in the
E. of the Murray Basin. The period is Post-
Blanchetown Clay (Upper Pliocene?) andPre-Rufus Formation (Middle? Pleistocene).
This was a high energy activity, as also was the
deposition of the early valley fill—the coarse
sand and gravels of the Monoman Formation.The typical surface expression of the
Coonambidgal Formation in the Dam area is
shown in Fig. 39. On Berribee Station (Fig.
4) there is a gentle slope to the river tract in-
stead of the usual cliffs. The dunes of the
Woorincn Formation are higher on the slope
than the Coonambidgal riverine sediments,
which occupy the bottom of the valley (Pis.
3-4). No interpenetration was found. TheCoonambidgal sediments occupy a younger part
of the river tract. This is in keeping with the
radiocarbon datings. Carbonate in the surface
soil of a Woorinen dune on Berribee Station
dated 14,200 + 790 y. (N.Z.) Fossil woodsat the base of the Coonambidgal Formation (asinterpreted in this paper) dated 4,040 and7,200 y. respectively. On Berribee Station theLindsay River forms an anabranch with theMurray River, and the enclosed land is called
Lindsay Is. The collagen of Aboriginal bonesfrom a burial in the top of a Coonambidgalchannel border dune on Lindsay Island dated
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 49
3,580-± 370 y. (ANU-420D). Bowler (1967)
dated Coonambidgal sediments as 5,110 ±130 y. (N-302). A closer definition of the
Coonambidgal Formation and more dates are
required to define the period of riverine activity
that its sediments represent.
(d) Stratigraphic Correlation in Murray Basin
The stratigraphy and geomorphology of the
study area described above present a broad
terrain consisting of Blanchetown Clay and as-
sociated deposits overlying marine beds. On this
is a veneer of mostly windblown deposits. Due
to the limited supply of suitable sediments, this
veneer is relatively thin and in some places
absent. Thus it is a common sight to see water
reservoirs (tanks) cut in green Blanchetown
Clay. Into this terrain the Murray River incised
a deep tract, now partly infilled with the Mono-
man and Coonambidgal deposits. However,
this W. side of the Murray Basin contrasts with
the E. side. The latter has no marine beds in the
stratigraphic succession, but a thick complex
of mostly riverine deposits, often coarse, that
contrast with the usually fine riverine sediments
capped by the extensive lacustrine Blanchetown
Clay with lenses of dolomite and dolomitic
limestone found on the W. side of the Murray
Basin. The terrain is a relict one (Pels 1966)
and consists of the channel and floodplain
deposits of ancestral and prior streams (Bowler
1967, Bowler and Harford 1963, 1966, Bowler
andMacumber 1968, Dury 1963, Butler 1950,
1958, 1960, 1961,Langford-Smith 1960, 1962,
Pels 1964a, b, 1966, Stannard 1962). Both
tectonics and climatic change were master
factors in the process. Earth movements of the
Bass Strait Epoch began in the Mesozoic (Fig.
3) and faded in the lower Tertiary. Those of the
Kosciusko Epoch began at the end of the
Miocene and there are small displacements in
Bookpurnong Beds in the Murray Basin, but
the main uplift was in Upper Pliocene and
Lower Pleistocene times. Tectonic movements
caused diversion of streams over a long period
(e.g. Pels 1966). The uplift of the Dividing
Range resulted in two processes significant for
the present study:
1. The down-warping of the country inland
from the Divide (E. side, Murray Basin)
so that riverine sediments 1500 km from
the sea extend below sealevel.
2. The construction of a massive bahada or
apron along the inner side of the range by
the formation of numerous extensive coales-
cing fans.
It has not been evident before what became
of these active prior and ancestral streams.
The stratigraphy described shows that, follow-
ing the retreat of the sea, they faded out in the
W. into low energy streams, and into the series
of lakes that deposited the Blanchetown Clay.
With each phase of the downwarp, the rivers
graded to a new base level.
On the Darling River NNE. of the study area,
in the vicinity of Menindee, is a string of lakes
that under natural conditions is dry except in
flood time, but are now used for water storage.
Connecting these lakes is a gypsiferous greenish
gray clay that I think should be referred to the
Blanchetown Clay. Thus in this area also there
are extensive fine lacustrine sediments forming
part of the lake system or systems into which
the ancient streams flowed from probably the
MURRAY BASIN
L'icmtrme &. riverine deposits Oulwash sedimenti
GREAT DIVIDING
RANCH
GREAT DIVIDING
RANGE
MURRAY BASIN
Fig. 41—Notional diagram (in section and plan) of
the relationships in the Murray Basin of the
oppositional fans along the foot of the
Dividing Range to the fine lake/river sedi-
ments described in this Memoir. The LakeMenindee and Lake Bungunnia systems are
separated because of their significantly dif-
ferent levels at the present time, but the
matter has not been investigated.
50 EDMUND D. GILL
Upper Pliocene to the Lower Pleistocene (see
section on Chronology). Significantly, this wasthe period of active tectonic movements whenthe Pinnaroo Block was uplifted and extensive
damming of the Murray occurred. Figure 41
is a schematic representation of the concept out-
lined above.
(e) Lake Victoria Sand
This stratigraphic name is hereby proposed
for the large and complex lunette (in the
original sense of Hills 1940 for a crescentic
dune) on the E. side of Lake Victoria. It is
composed of three members:
3. Dunedin Park Sand. Mostly 0-200 yr
2. Talgarry Sand. Upper Holocene
1. Nulla Nulla Sand. Upper Pleistocene to
Lower Holocene.
The Lake Victoria Sand is named after the
lake it borders, while the three members are
named after the three pastoral stations that
include areas of the lunette on their properties.
The low dynamics of this flatland river
system have already been discussed. A result
is that even the dune sands are finer than is
normal, as can be seen in the section on grain
size analyses. Another unusual feature of the
lunette is the comparative rarity of normal dunestructures due to sandfalls. The majority of
the bedding is subhorizontal (PI. 10, fig. 1).
Gray eolumnar soil
The inference is that the wind dynamics ex-
ceeded the sand supply, so that the dunes during
building were in a more or less constant state
of blowout. The smaller grain size of the sand
meant that it was lighter and could more easily
be blown.
The lunette is a large structure 15 km long
and up to 2 km wide, though more usually 4-
8 km in width. It thins out at each end, but
the two members thin out differentially. TheNulla Nulla Sand thins earlier towards the S.
end, while the Talgarry thins out earlier towards
the N. This structure could be due to small
differences in prevailing winds at the time of
construction (considered more likely) and/ordifferences in the river regime supplying the
sand. The S. end of the lunette is tied to anoutlier of Rufus Formation. The lake side of
the lunette is characterized by a frill of gulches
and interfluves (Fig. 16) while the inland side
is notable for its numerous sandfalls of DunedinPark Sand. The crest of the lunette is commonlyblown out leaving mesa-shaped residuals of
Talgarry Sand, with its characteristic gray
columnar soil (PL 10, fig. 1). The floor of ablowout is usually a subhorizontal paleosol.
The result is a long flat blowout across the
lunette with a sandfall at the inland end. Alongthe lower part of the lake slopes there is a sandapron (14° steepest slope where measured),compounded of sand blown up from the beachby W. winds, and fans of sand washed out of the
Gray columnar soil
Gray Paleosol
TALGARRY SAM) MEMBER
Red Paleosol
Red Palcosuls
NULLA NULLA SAND MEMBER
Red Paleosol
Fig. 42—Sketch of S. wall of Gulch N12 (Fig. 16), in the lunette on the E. side of LakeVictoria, as revealed by drought conditions in 1968. This is the largest number ofsuperposed paleosol units seen in the lunette. The basic occurrence is shown inFig. 4 of Gill on Palaeopedology, this Memoir. Site 8
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 51
gulches by the infrequent rains. The apron is
included in the Dunedin Park Sand.In 1967-8 during a drought, the stratigraphy
of the lunette was revealed by wind erosion andabsence of vegetation. Difficulty was experi-enced in distinguishing the numerous gulches,so those N. of the Nulla/Talgarry fence havebeen numbered northwards using the prefix N.,while those S. of the fence have been given theprefix S. (Fig. 16). Interfluves are given thusNl/2. All the gulches N. of the fence and manyof these S. of it were marked at their headsby long pegs with numbers drilled into thewood, as paint would be sand-blasted away.In addition, four deeply-sunk concrete benchmarks were inserted in the area of most intensestudy. These are linked to the Chowilla Damsurvey (and so sealevel at Adelaide), asfollows
:
47 m above MSL AdelaideBM 1 47 mBM 2 43 mBM 3 49 mBM 4 47 mThe foregoing system makes it possible to
locate a site with accuracy, and give its R.L.Because of the subhorizontal structure of the
lunette, such an R.L. is more significant thanin most dune systems.
Fig. 42 is a sketch of the N. wall of the inter-
fluve N14/15. This section shows clearly the
two main Members—Nulla and Talgarry. Thedistinction is not always so obvious, because in
such an environment there is always some sand
movement and remodelling complicating the
stratigraphy. Overall, the two Members are
readily distinguished, and are important in that
they represent two periods of dune building.
The paleosols represented in Fig. 42 constitute
the largest number seen in one section. Such
paleosols reflect the general alternation ofstable and unstable conditions seen in thein the E-W. dunes. By means of radiocarbondating it would be possible to make correlationswithin the range of the method. The few assaysit was possible to make are given in the sectionon chronology, where the possibilities of suchan investigation can be glimpsed.
The two Members making up the bulk ofthe lunette can be distinguished as follows:
Nulla Nulla Sand
1. More compact (3-
5 kg/cm2)
2. More leached
Free carbonate rare
(unless transferred
from above)
Selenite needles rare
3. Red non-columnarpaleosols
4. Internal structures re-
duced5
.
Extinct marsupials
6. Human remains only
in top
7. Fossil bones mineral-
ized (may be coated
with carbonate)
Clay more organized
as skins
Talgarry Sand
Less compact (1-
3 kg/cm2)
Less leached
Free carbonate
commonRhizomorphs pre-
sent
Selenite needles
commonGray columnar
paleosols
Internal structures
fresh
Extant marsupials
Human remains
throughout
Fossil bones not
mineralized (or
much less than
those in the
Nulla Sand)Clay mostly free.
There is much variation, the usual difficulties
of dune stratigraphy being present. However,the differences are real. A check with a Munsellchart showed that, although occasionally the
Dunedin Parte Sand
W UIk \'ictor
LAKE \ K lOKIA SAM) FORMATION
TALGARRY SAM) MEMBER
'Red PuL-omiI-
iy columnar soil
y-^c" Sandhills at Dunedin Park Sunt!
Fig. 43—Diagrammatic E-W. section of the Lake Victoria lunette to show the subhorizontalnature of its essential structures. The summit of the lunette is wind-eroded, resultingin the Dunedin Park sandfalls to the E. The hamada of Dunedin Park sand on thelake frontage is partly blown up from the beach and partly washed down from thegulches. The prevailing wind is W.
EDMUND D. GILL
lowest Talgarry soil becomes reddish, it does
not become as red as the Nulla Nulla paleosols.
The fossils are an important difference.
During the depositional break between the
two members, blowout structures developed. In
places the paleosols of the Nulla Nulla Sand are
truncated on the W. face of the lunette (Fig.
43). In NS. section deep gulches are seen in
the Nulla Nulla Sand that have been infilled
with Talgarry Sand (e.g. N14).The erosion of the red paleosol in the
fossil swale is part of the evidence of this pro-
cess. No such corrasion has been noted during
the emplacement of the Nulla Sand. Thus, in
the late Pleistocene, when the giant marsupials
were becoming extinct, deposition gradually
ceased on the Lake Victoria lunette. The time
of maximal world glaciation of c. 18,000 yr
B.P. is significant for terrace building and re-
gional (E.-W.) dune building, so perhaps it
was also for this lunette in the river tract.
Because of the erosion between the two mem-bers of the Lake Victoria Sand, the top of the
Nulla Sand has different ages in different sites,
e.g.:
1. Mussel shells from midden in red paleosol
a short distance N. of the Nulla/Talgarry
fence where the Nulla Sand is uneroded,
6,360 ± 140 yr B.P.
2. Mussel shells from midden near peg N18at head of gulch 5 3m below top of Nulla
Sand, 15,900 ± 275 yr B.P.
3. Mussel shells in Nulla Sand at N. end of
lunette 1 2 km SE. of Old Hotel site, Nulla
Station, 16,720 ± 260 yr B.P. Charcoal
from this midden 15,300 ± 500 yr B.P.
4. Mussel shells from midden in red paleosol
in second last gulch at S. end of lunette,
Talgarry Station (Fig. 44), 17,530 ± 320
yr B.P.
Corrasion tends to stop at a paleosol be-
cause such are much more difficult to erode. The
foregoing evidence suggests that the erosion
interval was mid-Holocene, the same time as
the interval between the Monoman and
Coonambidgal Formations. However, more
work on the stratigraphy and many more dates
are required to establish this. On Moorna
Station, on the high ground above Triple
Swamp at Triple Swamp Gulch (PI. 9), an
Aboriginal midden was found on the B horizon
of the soil where the original A had been eroded
(apparently by deflation) and then recon-
stituted. The mussel shells from between the
original B and the re-built A horizon dated
7210 =±= 160 yr B.P. This stripping could
therefore belong to the same period.
Thus, on present evidence, the Nulla Sand
was built up mostly in the late Pleistocene. Theincluded paleosols dated by the middens in
them (14,000-17,000 yr) appear to tie up with
the paleosols of that range of CI 4 dates in the
E-W. dunes on Berribee Station, at Ouyen and
at Swan Hill, even with soil formation in S.
Victoria (see section in chronology).
A lakeside profile was surveyed (Fig. 45)when the lake level was at its lowest during the
drought. A wide beach of white sand was ex-
posed, and such is the source for sand blown
up the profile by the prevailing W. winds.
One may well ask what becomes of all the
clay transported by the river with this sand.
Much clay has accumulated on the floor of
Lake Victoria. The Darling River is muddier
Gray columnar seal
Gray Paleosol
LUNETTELAKE VICTORIA SAND FORMATION
Gulch
NEW SOUTH WALES
Aboriginal shell midden
Dibconformtty ->^ 1
7
.530 yr C14
SITE 9
Talgarry Sm.
TALGARRY SANDMEMBER
boundary
NULLA NULLA SANDMEMBER
Fig. 44—Section of second last gulch at the S. end of the Lake Victoria lunette (Fig. 16).Site 9.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 53
than the Murray because its waters are fresher.
As the percentage of dissolved salts increases
(especially sodium), as in the Murray River,
the clay flocculates. Much is deposited in the
low energy anabranches and billabongs or
oxbows (e.g. Fig. 7). Such a stream does not
penetrate the ground easily or erode its banksby wetting and weakening. Erosion is by phy-sical processes. Much clay is blown up into the
lunette as aggregates or attached to sand grains.
In swales and such low areas, strong wetting canre-organize the clay as clay skins that bring
about the compaction of the sand. The fineness
of the sand and the evenness of grain size makesthis process the more effective. Numerous fossil
claypans of this type can be seen in the Nulla
Sand but they are rare in the Talgarry Sand.
They erode with a typical fine herringbone
pattern. Sections of such pans can be seen in
the walls of gulches.
River and wind action have thus combinedto reduce clay content and build a stockpile
of fine sand over many millenia. This lunette
sandpile has been roughly estimated at
100,000,000 tonnes. It has provided a habitat
for marsupials, placentals, reptiles, birds and
man. This biologic history has been preserved
because of the readiness with which the dead
animals were interred in the rather dry sands.
However, the river system is no longer ade-
quate to transport sand to this site, but the
wind energy is adequate (with the aid of in-
troduced animals) to erode the dunes, revealing
their inner structure and fossil content (Mar-
shall, this Memoir).
(f) Lake Victoria Hamada
No formational name has been created for
the colluvial fans that overlie the BlanchetownClay and associated formations that constitute
the scarp along the W. side of Lake Victoria.
To a certain extent they are the time equivalents
of the Lake Victoria Sand. The youngest
member is a series of recent fans that are ob-
viously the latest product of sediment flow
on these slopes. The fans are geomorpho-logically complete except for contemporarygullying. Although so young, the fans have a
surface held by carbonate precipitation, usually
within the range of 5-15 cm deep. This earthy
carbonate has preserved Aboriginal skeletons
and middens. In the field we called it the "in-
durated layer", and it can be traced across all
the erosion amphitheatres (Fig. 14) on the W.side of the lake. It continues right up to the
edge of the general terrain some 35 m above the
lake. The hamada slope is of the order of 10°.
About 400 m N. of a windmill and J. May'scamp (shown in Fig. 17) on Noola Station is a
typical fan which includes Aboriginal middens.
Associated with one of these were carbonized
twigs that looked like salt bush. A radiocarbon
assay gave a "modern" age (ANU-423),which is less than 200 years. As the middenwas immediately below and affected by the
indurated layer, this accumulation of carbonate
is very recent, and the gullying even morerecent, i.e. during European occupation. Thewidespread occurrence of the indurated
layer shows that the hamada had achieved a
certain degree of stability. Figure 46 shows a
LINETTL LAKE VICTORIA SAM> 1 ORMATION
I \L(i \KRY SAND /""""TV OwQ ioliimnar soil
MEMBER £1^
Red Pakua
NULLA MELASAND MEMBER
Nulla S(n.
Ffg. 45—Sketch of an actual lakeside profile in the Lake Victoria
lunette, Nulla Station, N.S.W. Site 8. Boundaries be-
tween the Nulla Nulla Sand and Talgarry Sand are
subhorizontal normal to the lake, but parallel to the
lake are scalloped in places by fossil gulches. The struc-
tures reveal successive phases of stability of the terrain.
54 EDMUND D. GILL
NOOLAEROSIONAMPHITHEATRE
Greenish vr.w
BLANC IIlYownCLAY
Mesa formed by
indurated layer
ol uarbonate
Noola Sin.
M_\V SOUTHWALES
Promonton k-vel Inrincr lake [five
Mli 6
Fig. 46—Sketch of slope in Noola Erosion Amphi-theatre showing the latest pedocal, consist-ing of indurated layer 10-15 cm thick,
which covers the low fan slopes (c. 10°)at the foot of the scarp forming the W.shore of Lake Victoria, N.S.W. Carbonizedtwigs from this layer dated 'modern' byCI 4, so the soil is less than 200 yr old.
The strong dissection has therefore occurredsince European occupation. Aboriginalbones and middens arc cemented in thislayer.
typical dissection of the indurated layer on theN. side of the promontory constituting the S.
limit of the Noola erosion amphitheatre (Fig.
14). This layer is part of a 1 -5 m oxidized zoneon the promontory surface (
0°-3 °) , below
which is a similar thickness of mottling in the
Blanchetown Clay. On the 10° slope from this
promontory is 15-25 cm of indurated layer
resting directly on 1-5 m of mottled Blanche-town Clay. In the middle of the Lake Victoria
amphitheatre the oxidized zone attains a thick-
ness of 4-5 m,The development of a hamada that was later
dissected is a process that has occurred a num-ber of times on this lake scarp. It is part of
the impermanence of semi-arid terrains. Figure
47 shows a fossil gulch which contained bettong
bones. The bones were strewn out in a mannersuch as occurs in hillwash, so the animal wasnot burrowing, but has the same age as the
sediments.
A gulch 11m deep near the N. wall of the
Lake Victoria amphitheatre, about 100 m W.of a mesa reveals a Pleistocene gulch fill of light
brown (7*5 YR 6/4) clayey sand to brown
(5/4) sandy clay with small lenticles of gravel
(Fig. 48). A zone of earthy carbonate 25-
40 cm deep occupies the surface of the fill.
A small Aboriginal midden and fireplace of
oval section was present 45 cm from the sur-
face. It consisted of well compacted (8-10
kg/cm-) ash, charcoal and mussel shells 20cm deep and 45 cm wide. The general colour
was brown (7 5 YR 4/2 to dark brown (3/2).The bottom of the structure was lined with
charcoal suggesting an excavated fireplace.
Many of the mussel shells were on edge, stand-
ing vertically. Four worm burrows were visible
in vertical section and two in horizontal sec-
tion. None could be found in the surrounding
sediments. This is taken to mean that a zoneof high organic content attracted the worms.A radiocarbon dating of the charcoal gave
18,200 ± 800 yr B.P. This is the oldest
Aboriginal site dated by us in the project area.
Chronology
"The same regions do not remain alwayssea or always land, but all change their condi-
tion in the course of time". —Aristotle (384-322 B.C.)
1 . Methods
The chronology of the formations describedis difficult because no materials suitable for
isotopic assay older than the range of radio-
carbon have been found. Moreover, the forma-tions exposed in natural sections of the countryrock over most of the area are Moorna Forma-
Clavey sand infilling former gulch Auger Shallow gulch
Fossil bettoag bones in carbonate soil
Fig. 47—Measured section of gulch (10-4 mwide, 1-2 m deep) in hamada on W.shore of Lake Victoria, cut in the fill ofa still older gulch. The S. bank shows38 cm of brown 7-5YR 5/4 clayeysand over an earthy carbonate zone (Bhorizon) 69 cm thick. Nearby this zonewas dated 18,200 yr (GaK-2514) on thecharcoal of an Aboriginal midden. Inthe above section bettong bones Weretaken from the carbonate cement.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 55
tion, Blanchetown Clay and Chowilla Sand.These sediments suffer lateral differentiation
and considerable interdigitation, thus com-plicating correlation. The chronology is there-
fore based on certain datum planes that traverse
the area, as follows:
5. Chronometric dating by radiocarbon.
4. Paleosol that mobilized carbonates (cal-
crete ) and beyond the range of C 1
4
(earliest pedocal).
3. Paleosol that mobilized silica (silcrete, com-mon opal, disperse silica giving silicihed
sands).
2. Paleosol that mobilized non-magnetic iron
oxide (laterite).
1. Bed with Cheltenhamian marine fossils, metin oil bores and in the Tareena water well
(Tate 1899).
These stratigraphic planes can be demon-strated to be superposed.
The paleosols provide datum planes that arc
independent of the lateral differentiation andinterdigitation. They also define successive
changes of climate and related geochemicalconditions. This palaeoclimatic sequence canbe paralleled elsewhere, which assists extrapola-
tion. Further, Marshall (this Memoir) hasidentified in the Moorna Formation vertebrate
taxa comparable with those found in Pliocene
beds elsewhere. This formation lies below andis older than the Blanchetown Clay andChowilla Sand. In the valley of the MurrayRiver an oxidized lluviatilc terrace (Rufus
Formation) contains a vertebrate fauna with
taxa such as occur in Pleistocene deposits else-
where. Vertebrates also assist with the chron-
ology of the lunette on the E. side of LakeVictoria, where the lower part of the formation
contains large extinct marsupials, while the
higher part contains only extant taxa.
An attempt was made to use palynology,even though it has a restricted application in
this dry, oxidized terrain. Two horizons ofdark gray to black fine sediments were found,one at the top of the Parilla Sand in the MurrayRiver cliffs beside Kulcurna Homestead, andanother nearby at the S. end of Salt Creek in
the Blanchetown Clay where it thins out onan old lake shore. The former provided nodefinitive evidence, but the latter gave a flora
more or less identical with that of the present
semi-arid terrain (Churchill, this Memoir).There is thus no chronologic value in the results,
but they are of profound ecologic interest in
that there is an indication of the time of incom-ing of the present aridity following the sub-tropical rainforest milieu evidenced by the
laterite. This aspect is dealt with in the sec-
tion on palaeoelimatology. Every method that
appeared likely to assist the chronology wasattempted, including fission track dating of
opal, palaeomagnetism, thcrmoluminescence,neutron activation, the U/Th assay of pedo-genic calcretes, and fluorine/phosphate analyses
of bones. Reports on the application of some of
these methods are given in accompanyingpapers. By synthesizing the results of the datumplane study and the above methods with the
tectonic and palaeoclimatic patterns, a chron-ology has been derived which is about as far
as one can go under present circumstances.
2. Tectonic Pattern
In the S. Australian River Murray region the
late Cainozoic stratification is regular, and con-sistent over a great area. The sequence is;
Youngest Bungunnia Limestone
Blanchetown Clay
Chowilla SandOldest Parilla Sand.
Null:! Stn.
sn u 7
LAKH VIC IOKIA
Aboriginal midden under ridge
CI4 Samplfi| (5,300 yri
bcucli
Ki-.sklu.il off section lino
IKIl.ilal .iK.i
Sand ridjja
Lakt Victoria hiyh vwicei level LAKE VICTORIA SANDS 1 OKMAl ION Ml LA Nl'l 1 A SAND MI.MIU K
Fig. 48—Surveyed section 292 m long on the N. shore of Lake Victoria E. of the Old Hotel sile at
end of track from Nulla Station.
56 KDMUND D. GILL
As can be seen by Fig. 18, the Parilla Sand
Outcrops strongly at Ihe dam site in the valley
walls. This is true also as far east as Kulcurna
Homestead. Ihe formation is then affected bythe lake Victoria Synciine and disappears from
sight in the river banks, not to rc-appear again
till Paringi 29 km SE. of Mildura. About 15km SE. of the Paringi turnofT the MurrayRiver dill's come close to the Robinvale-Gol
Ciol road (Sturt Highway). There cliffs about
25 m high show 4 5 m of Parilla Sand which,unlike any other formation present, develops a
vertical face. Hie distance between these twooutcrops oi' Parilla Sand is about 130 km in a
direct line. The Parilla Sand is a gray fluviatile
clayey sand of constant character, apparently
not affected by significant tectonic movementsduring its emplacement. It has since been Hexeddownwards in the Lake Victoria Synciine. Bycontrast, the succeeding formations have beenaffected by tectonic movements. The MoornaFormation, interposed between the Parilla
Sand and overlying formations, is limited (as far
as we know) to the Lake Victoria Synciine.
Above the Moorna Formation is the Blanche-town Clay/Chowilla Sand complex. In S.
Australia (on the uplift block) these formationshold a definite order, but in the depressed LakeVictoria Synciine area, they grossly interdigitate
and the Blanchctown Clay reaches a muchgreater thickness (PI. 5). The onset of signifi-
cant Kosciusko Epoch earth movements canthus be recognized in the study area. TheMurray Basin is so large a structure, that tofit its earth movements into a tectonic pattern,
one needs to view the tectonics of the whole of
SE. Australia and not movements on local faults
as at Melbourne and Adelaide. This general
view shows a mid-Tertiary quiescence withslight movements at the end of the Miocene andin the Lower Pliocene, then strong movementsin the Upper Pliocene and Lower Pleistocene.
Since the movements in the Lake Victoria
Synciine are so subdued (deep movements are
buffered by L5-3 km of soft sediments—Fig.
3), it is reasonable to conclude that the well-marked changes following the deposition of theParilla Sand are those o\' the Upper Pliocene.
If this be so, then the Moorna Formation re-
presents the beginning of the damming of the
Murray River by the uplift of the Pinnaroo
Block, while the lacustrine deposits of the
Blanchctown Clay (and associated limestone
lenses) mark the establishment of this imped-
ance to drainage. It would seem that this wasvariable, because after some lacustrine deposits
were emplaced, the lake dried up so as to yield
dolomite then pedogenic opal (Fig. 26) andsilcretc. The oscillation between channel, flood-
plain and lacustrine facies, as reflected in the
inierdigitation of the Chowilla Sand/Blanche-town Clay sediments, bears witness to these
changing conditions.
This argument from tectonic pattern makesthe Moorna Formation Upper Pliocene (a con-clusion reached by Marshall in this volume onthe evidence of vertebrate fossils) and the
Blanchctown Clay/Chowilla Sand complexUpper Pliocene to Lower Pleistocene, if we candefine what that means. This will be attemptedin the next section which deals with the Plio-
Pleistocene boundary. Fig. 3 shows post-Eocenesedimentation to have been regular until post-
Bookpurnong times, after which an appreciable
thickening of sediments occurred in the LakeVictoria region. The Bookpurnong Beds areuppermost Miocene in the study area as judgedby the fauna from the Tarecna well, alreadydiscussed. However, it should be noted that
this formation over its very wide distribution is
somewhat diachronous (Lawrence 1966, p.
534), apparently due to the slow retreat of thesea in the Murray Basin from E. to W. Theargument from the tectonic pattern is thusconsistent with the other methods of dating.
3. Plio-Pleistocene Boundary
Discussions about which beds are Plioceneand which Pleistocene have little meaning underpresent circumstances, unless one states what is
accepted as the boundary. Shotton (1967)and others have pointed out that various in-
terpretations of the boundary fall between 1 75and 3 5 m.y. A review of the Plio-Pleistocene
boundary in Australia has recently been made(Gill 1972b).
In 1948, INQUA defined this boundary byunanimous vote, so the procedural position is
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 57
unequivocal. The chronometric dating and ex-
trapolation of this horizon is the problem. In
the writer's opinion, the best approximation onpresent evidence is the dating of the boundaryat about 18 m.y. This is used temporarily in
order to make objective what is meant in this
paper by the terms Pliocene and Pleistocene.
In the section on palaeoclimatology the
changes in climate are described. These have
general chronologic significance, as also do the
changes in flora that accompanied them. Im-portant changes in fauna also occurred, but
these have yet to be elucidated before they can
be used for definitive dating. Changes in soil
type also accompanied changes in climate, and
these are described in the accompanying paper
on palaeopedology.
4. Radiocarbon Dating
The following dates were obtained from
assays of materials from the study area:
1. Modern (ANU-423). Wood charcoal from small
shoots in surficial layer indurated with secondary
carbonate. Associated with an Aboriginal midden of
mussel shells (Velesunio ambiguus). The site is on a
fan forming part of the hamada on the NW. side of
Lake Victoria, about 0.4 km N. of J. May's campnear the lake windmill on Noola Station, N.S.W.,
Anabranch Military map ref. 422,800. The result
dates Aboriginal occupation of the site, provides amaximal date for the latest lithification by secondary
carbonate, and proves the modern age of the wide-
spread gullying that cuts the indurated layer.
2. 180 ± 80 yr B.P. (GaK-2511). Charcoal frommidden of Velesunio shells on the N. side of the
polythene water pipeline from near the Talgarry
Station homestead to Lake Victoria, N.S.W. Thesite is on high ground (near top of ridge with
Casuarina copse) overlooking the salt lake behind
the lunette (Field sample CHA/253). Dates Aboriginal
occupation, and in a general way the uncompactedsandy soil in which it occurred.
3. 770 ± 150 yr B.P. (GaK-3217). Finely dis-
seminated carbon from core taken in floor of MoornaEast oxbow (billabong) below soft loose sediments
and overlying a highly compact mottled zone (B
horizon of paleosol) of pre-oxbow land surface. Col-
lected 23 May 1969. Recovered carbon only 0-7 gmso the date could contain an unexpected error. How-ever, the date is in keeping with the recent age of the
oxbow inferred from sharp banks and low talus
slopes. Moorna Station, W. of Wentworth, N.S.W.4. 1320 ± 80 yr B.P. (GaK-2008). Charcoal fromlarge Aboriginal midden of Velesunio shells beside a
prominent track S. of the pipeline on Talgarry Station
E
on a low dune associated with the inland limit of the
lunette complex, W. of the salt lake and fenceline.
Collected 22 April 1968. Illustrated in Gill 1969, p.
177, fig. 2.
5. 1660 + 110 yr B.P. (Gak-2512). Charcoal fromAboriginal midden on the S. side of the pipeline
track mentioned under (2) but E. of Casuarina copseat top of slope, Talgarry Station, N.S.W. In uncom-pacted sandy soil, SW. of dam (at gate in fence).
6. 1930 ±80 yr B.P. (GaK-2007). Charcoal fromAboriginal oven with calcrete ovenstones in sandon top of hill near Salt Creek SE. of Dickie's Gatebeside copse of trees in NE. corner of Scrub Paddock,Kulcurna Station, N.S.W. Collected 28 April 1968.
7. 3580 ±370 yr B.P. (ANU-420D). Collagen ofAboriginal bones from burial in top of channelborder dune, Lindsay Island, Berribee Station (Fig. 5),Victoria. Sample CHA/174, site 2.
8. 4170 ±200 yr B.P. (GaK-1432). Collagen ofAboriginal bones (extended burial) from Site 16, Ly-bra Paddock, Keera Station. W. of Merbein, Victoria
(Mildura Military Map ref. 481,779). In SE. Aus-tralia collagen is leached away after 5-6,000 yr so that
bone datings are seldom possible on this fraction if
beyond that age. The four samples from this site
(GaK- 1430-3) are intrusive burials, and althoughreasonably sized samples were provided the yield wassub-standard. As a result GaK- 1431 was measuredunder low gas pressure and 1432-3 were diluted withdead carbon. The dates may therefore not be ac-curate, but it is noted that where two skeletons oc-
curred one above the other, the ages are in the
correct order, viz. 4170 and 4400 yr.
9. 4400 ±220 yr B.P. (GaK-1433). Same type ofmaterial and same site as foregoing sample. Ex-tended burial.
10. 5350 ±290 yr B.P. (GaK-1431). Ditto. Flexedburial.
11. 5840 ± 90 yr B.P. (GaK-1429). Charcoal closely
associated with loosely flexed Aboriginal skeleton,
Brown's Paddock (Fig. 38), Keera Station, Victoria.
Mildura Military Map ref. 469,777.
12. 5900 ±550 yr B.P. (GaK-1430). Collagen ofAboriginal bones from Site 5, Lybra Paddock, KeeraStation, Victoria. This was a sitting burial. Thesamples from Keera Station were selected to covervarious types of burial.
13. 6360 ± 140 yr B.P. (GaK-2416). Velesunio shells
from Aboriginal midden in red paleosol between theNulla Nulla Sand and Talgarry Sand (members ofthe Lake Victoria Sand formation) where no inter-
Member erosion has occurred. This is therefore aminimial date for the Nulla Nulla Sand. Site 50,Nulla Station, on lunette on E. side of Lake Victoria,
N.S.W.
14. 7200 ±140 yr B.P. (GaK-2513). Small log ofEucalyptus largiflorens (Black Box) determined byMr. H. D. Ingle, C.S.I.R.O., on wood structure. Theengineer overseeing the test trench for a bituminousgroundwater cut-off curtain at the Chowilla Dam
58 EDMUND D. GILL
site (S. Aust.) showed me the heaps of sedimentsamples from various depths that he prepared for con-tractors. The fossil wood came from the samplecovering 7-5-10-5 m from the surface (25-35 ft). Thelog was about 10 cm in diameter so that there is asmall factor of biologic age.
15. 7210 ± 160 yr B.P. (GaK-1726). Velesunio shells
from Aboriginal midden between the A and Bhorizons of surface red (5 YR 6/4) soil on E. side
of Triple Swamp Gulch, Moorna Station, N.S.W.The A horizon has therefore suffered stripping, thenbeen replaced. This history is reflected in the lack
of compaction in the A horizon, viz. 2-4 kg/cm2 (Bhorizon 4-8-8 kg/em2 but harder in zones of second-ary carbonate deposition). The shells were excavatedby Sir Robert Blackwood from 0-5 m3 of sandy soil
during the survey of the gulch (Fig. 24).
16. 9160 + 340 yr B.P. (GaK-2921). Bone apatite
from bones of extinct marsupial collected 30 Oct.
1968 by H. E. Wilkinson from Nulla Nulla Sand in
lunette on Nulls Station, E. side Lake Victoria, N.S.W.The sample of about 700 gm included an imperfecttibia, scapula and metatarsal of Protemnodon. Boneshad a carbonate encrustation such as is not seen onbones from the Talgarry Sand. Site 50, from blow in
interfluve 4-5 N. In view of the current discussions
on apatite datings, this result should not be taken tooliterally. However, it is consistent in that it falls be-tween the limiting dates for the Nulla Nulla Sand of6,360 and 17,530.
17. 11,250 ± 240 yr B.P. (GaK-1062). Selected thick
pieces of well-preserved Velesunio shell from base oflarge Aboriginal midden on top of high cliff on left
bank of Murray River at Redcliffs, Victoria (34° 19'S,
142° 17'E). The cliff sections a high terrace. Theantiquity of the midden was originally surmised fromthe degree of compaction, and that the midden wassectioned by the high cliff.
18. 12,600 ±1300 yr B.P. (ANU-404B). Charcoalfrom Aboriginal mussel shell midden associated withsample 404A (dated 24) which dated 17,530 yr.
Sample was only 1% of requirement.
19. 14,200 ±790 yr B.P. (NZ R2729/1). Earthycarbonate from 0-6-0-9 m in auger hole (No. 2) sunkthrough top of an E-W. dune of the Woorinen Forma-tion c.l km SSW. of Berribee Station homestead(Fig. 10), NW. Victoria. (Sample CHG 88/33). Dr.T. A. Rafter, D.S.I.R. Institute of Nuclear Sciences,advised that the concentration of COa=10 5%, and$t\2~Z-(> i oo.
20. 15,300 ± 500 yr B.P. (GaK-2515). Charcoal fromlarge Aboriginal midden in Nulla Nulla Sand contain-ing also Velesunio shells (date 23), bones includingOnchogalea frenata, a gastrolith and a quartzite scra-per. The site is near the end of the track shown on theAna Branch Military map that finishes at the N. shoreof Lake Victoria. The ruins of an hotel stand there,
and the midden is c.l km E. of the old hotel, NullaStation, N.S.W. The midden was still partially coveredby compact Nulla Nulla Sand as seen in Fig. 48. The
midden measured 13 x 6 m and was very compact. It
was 15 cm thick and stratified with lighter and darker
layers. Site 52, sample CHA/248. Some burnt bone
was present. Bones are rare in shell middens as the
vertebrates were usually cooked in the ovens re-
presented now by small heaps of burnt stones with
charcoal. This may be the reason why bones of a small
animal only were found in this site. The midden site
was surveyed to a Chowilla Dam Survey bench mark,which had a compass bearing of 148° therefrom, was100 m (328 ft) distant, and 41 m lower. This is the
N. limit of the lunette, which cuts out at the ridge
traversed by the track to the old hotel. W. of that is
a large erosion amphitheatre which we called the
Old Hotel Amphitheatre.
21. 15,900 + 275 yr B.P. (ANU-405), Shells ofVelesunio from an Aboriginal midden in a stratified
context in the lunette on the E. side of Lake Victoria,
N.S.W. The midden is at the head of gulch 18 onNulla Station near the marker peg N18. A small mesaof Nulla Nulla Sand stands over the midden to adepth of 4-55 m. The flat top is due to a paleosol.
The sample (CHA/330) was excavated from the mid-den. Human bones were found at the outcrop of the
midden, and it appeared that they could come fromnowhere else as they were in among the loose shells,
but they were not in situ.
450
22. 16,400 ± 560 yr B.P. (GaK-3218). Earthy butsolid carbonate (i.e. sand lithified with carbonate) fromsection of E-W. dune exposed on the W. roadcut ofthe Calder Highway (Melbourne to Mildura), 3-2 kmN. of Ouyen, Victoria. Sample from B horizon ofpaleosol 76-94 cm from surface. Another paleosol0-7 m lower in the dune was not dated (see morpho-logy section re dunes). Collected 15 May 1969.23. 16,720 ± 260 yr B.P. (ANU-422). Velesunioshells from Aboriginal midden c.l km E. of old hotelsite on N. shore of Lake Victoria. See date 20 abovefor charcoal from same site, and discussion of chron-ology. Sample CHA/248.24. 17,530 ± 320 yr B.P. (ANU-404A). Velesunioshells from second last gulch in lunette on SE. shoreof Lake Victoria, Talgarry Station, N.S.W. Samplefrom red paleosol in Nulla Nulla Sand (Fig. 44) wherea blowout occurred later infilled with Talgarry Sand.These blowouts may have taken place during the Post-glacial thermal maximum (see date 13 above). SampleCHA/320. 5 Cu%6 00 ±20. When this date is
taken with No. 13, it can be seen that the top of theNulla Sand varies from 6,500 to 17,500 yr accordingto how deep erosion has cut into that Member.25. 18,200 ±800 yr B.P. (GaK-2514). Aboriginalmidden consisting of compacted fine charcoal withmussel shells in hollow in sand of gulch (Fig. 47) onNW. shore of Lake Victoria (Noola Station), N.S.W.,near fossil bettong bones. Sample CHA/289. Thisis the oldest midden dated in this area. With time thepieces of charcoal of a midden break down and be-come a compacted mass of fine carbon. After weaken-
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 59
ing by bacterial action (surfaces progressively retro-
graded by oxidation of the carbon to carbon dioxide),
the charcoal collapses and compacts as a result ofground pressure.
26. 27,500 ± 700 yr B.P. (N,Z. R2729/4). Soil car-
bonate from near carbonate nodule quarry between309 and 310 mileposts on E. side of Calder HighwayS. of Hattah and Mildura, Victoria. The road is cut
through an E-W. dune. An auger hole was sunkthrough the dune to obtain in situ samples of sedi-
ments and paleosols (see log and comment in section
on dunes under Geomorphology). X-ray showed the
mineral to be calcite, and that quartz was present.
27. 27,800 ± 1900 yr B.P. (GaK-1727). Outer 5 mmof large carbonate soil nodule (c.12 cm diam.) con-
sisting of laminated calcite. This zone encloses anolder generation of small nodules (date 30). Samplefrom bank at N. end of Moorna E. oxbow (Site 13),
Moorna Station, N.S.W. At this site, carbonatenodules are excavated for roads.
28. 28,000 ± 1800 yr B.P. (GaK-1728a). Inner part
(5-10 mm) of laminated cortex of soil carbonate
nodule from Moorna E. oxbow (date 27). The dif-
ference in figures between assays 27 and 28 does notnecessarily mean that there is an appreciable dif-
ference in age. See discussion in paper on palaeo-
pedology. The core of this same nodule was beyondthe range of the radiocarbon dating method used(date 30).
29. 31,300 ± 1,200 yr B.P. N.Z. R2729/2. Soil car-
bonate from auger hole 2 in E-W. dune on Berri-
bee Station, NW. Victoria (detail under date 19),
belonging to the Woorincn Formation. Depth 2-64-
317 m.
30. >31,700 yr B.P. (GaK-1728b). Small nodules
formed most of the soil carbonate nodule used for
assays 27, 28 and 30 from Moorna Station, N.S.W.
This sample from the centre of this compact calcific
nodule was beyond the assay range.
31. >34,300 yr B.P. N.Z. R2729/3. This is the third
soil met in the auger hole through the E-W. dune
on Berribee Station, Victoria, and is that occupying
the surface of the terrain before the dune was built.
See discussion of dunes in section on Geomorphology.
Comment: This is a small grid of dates for so
large an area, but more were not possible. The
dates are of course in "radiocarbon years"
pending corrections (such as that for the half
life of C14). The disparity between dates of
over 30,000 yr assayed on marine carbonate,
and dates on the samples given by the U/Thmethod, cause one to critically consider the
dates on terrestrial carbonate in this same time
range. Some kind of cross check is required. In
an attempt to do this, sets of samples from the
study area were sent to two laboratories with
different approaches to see if they could apply
successfully the U/Th methods described byKislitsina and Cherdyntsev (1967) and byHansen and Stout (1968). No positive results
have been obtained so far. Nevertheless, it mustbe significant that the many carbonate C14terrestrial dates known to me, all fall in the
correct order where there is stratigraphic super-
position. I introduced dating of terrestrial
pedologic carbonates in 1964 when working onthe geology of the Talgai Cranium (Darling
Downs, S. Queensland). At that time, car-
bonates were not usually believed to be of anyuse for radiocarbon dating, but as no other
materials were then available, it was decided to
try them out. The reasoning was that the bed-
rock had been leached by lateritization, andthat the carbonate in the shells and nodules
originated in Tertiary basalt in the headwaters
of the streams. This was taken into solution, so
releasing the original carbon dioxide. The mol-
lusca in the stream incorporated the CaO in
their shells, but using modern carbon dioxide.
Assay of modern mussels from the same river
system collected prior to atom bomb tests
proved this to be so (Institute of Nuclear
Sciences, N.Z.). Such shells in stream terraces
were used for dating. Fossil shells incorporated
in riverine sediments are dissolved by soil acids,
and the CaO redeposited in the B horizon as
nodules, utilizing soil air. Such nodules were the
second type of material used for dating.
In addition to the stratigraphic series, cen-
tres and outer layers of some nodules weredated. These also gave dates in the correct
order of age, and up to 2,000 yr difference wasfound between outer layer and core. Themineral was calcite, and petrologic examination
by C.S.I.R.O. (Dr. G. Baker) showed that nore-crystallization had occurred.
In putting perspective into events in the
study area, the radiocarbon dates elucidated
certain processes. For example, the dynamics
of the region are such that all the Quaternary
sands are of similar grade (see section on sedi-
mentation). The radiocarbon dates show that
with time there is an increase in compaction for
such sands when in similar environments.
5. Paleosols
The significance of these pedoderms for
60 EDMUND D. GILL
chronology and palaeoecology is discussed in an
accompanying paper.
6. Fluorine Test.
The Chief Chemist of the Division of Agri-
cultural Chemistry, Mr. J. O'Brien, kindly
arranged for the analysis of a series of fossil
bones collected during the Chowilla Project,
and the results are given in an accompanyingpaper by Mr P. J. Sinnott. For chronology the
Fluorine Indices (for definition see Gill 1954)and the percentage of nitrogen are significant.
These figures with stratigraphic detail and someinterpretation are set out below. The samplenumbers are as in the Sinnott report.
1. Piece of fossil fish from 0*9-1-2 m aboveChowilla Sand, in the erosion amphi-theatre S. of the homestead on NoolaStation, situated on the NW. side of LakeVictoria, N.S.W. Site 6.
2. Fragments from large marsupial pelvis in
Moorna Formation (or base of associated
Chowilla Sand, which were not distin-
guished at the time), Fisherman's Cliff,
Moorna Station, N.S.W. Site 13.
3. Bone from Chowilla Sand under Blanche-town Clay at Bone Gulch, Moorna Station,
N.S.W. Site 12.
4. Fragments of Procoptodon goliah fromRufus Formation on the property of Mr.J. Curtis, 'Karawinna', on Boy Creek, Lot22 Parish of Tulillah, Victoria (Fig. 4).
5. Fragments of large marsupial from RufusFormation between Frenchmans Creekand the Wentworth-Renmark road, Dune-din Park Station, N.S.W. Site 10.
6. Fragments of marsupial bone from inter-
fluve N14/15, lunette on E. side of LakeVictoria, Nulla Station, N.S.W. in NullaNulla Sand. Site 8.
7. Fragment of Procoptodon goliah fromNulla Nulla Sand, Talgarry Station, N.S.W.Site 9.
8. Piece of bettong bone from Aboriginalmidden in Nulla Nulla Sand, E. of OldHotel site, Nulla Station, N.S.W. Site 7.
9. Marsupial bone from Talgarry Sand, 9-1*2 m above base of formation at inter-
fluve 13/14N, Nulla Station, N.S.W. Site
8.
10. Fragments of human skeleton lightly
cemented with carbonate in the surficial
'indurated layer' of a hillwash fan in the
erosion amphitheatre SW. of the home-stead, Noola Station, NW side of LakeVictoria Site 6. Sample CHA/288.
11. Fragments of human skeleton (field num-ber 6/0) collected 26 Apr. 1967 fromLybra Paddock, Keera Station, N.W. Vic-
toria. CHA/35.
12. Fragments of human skeleton (field num-ber 1/B), Brown's Paddock, KeeraStation, N.W. Victoria. CHA/8.
13. Fragment of holotype of Zygomaturus vic-
toriae (S. Australian Museum P4986)from Lake Victoria Station at the time it
consisted of all the country in the LakeVictoria area. Mr. K. Crozier of *War-
rakoo' and Mr. D. Harvey of 'Nulla' kindly
checked the matter, and the well fromwhich the fossil came (depth '45-60 feet')
was probably to the north of the lake. Theformations of the riverine tract are not in-
volved, so it must have come from the
Blanchetown Clay, the Chowilla Sand or
the Moorna Formation. It did not comefrom the first formation, because it is
brown in colour and not whitish like the
bones found therein; also the bones are
undisturbed whereas those from the
Blanchetown Clay seen by us are distorted
due to movements of the clay (which con-tains montmorillonite). In the study areathe Chowilla Sand interdigitates freely withthe Blanchetown Clay. The Fluorine Testand the Nitrogen Test do not distinguish
between the Chowilla/Blanchetown andthe Moorna formations, so it cannot besaid (on this evidence) whether the holo-type comes from the Chowilla Sand or theMoorna Formation. However the Chow-illa Sand is always oxidized and theMoorna Formation usually not, so the fossil
(being brown) is more likely to belongto the Chowilla Sand. So far efforts tolocate the exact site have not been sue-
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 61
14.
15.
16.
cessful, but if they are, then the depth will
give some indication which formation is
involved.
Fragment of bone from National Museumof Victoria specimen P29500.
Fragment of bone from National Museumof Victoria specimen P28882.Fragment of bone from National Museum
C14
of Victoria specimen P29502.
Specimens 14-16 are from the MoornaFormation and associated Chowilla Sand,
Fishermans Cliff, Moorna Station, N.S.W.
Site 13.
The results of the assays according to for-
mations, and the radiocarbon dates where
available (for control), are:
Formation
A. Indurated layer
(no formational name)B. Talgarry SandC. Nulla Nulla Sand
D. Rufus Formation
E. Intrusive in Rufus Formation
F. Blanchetown ClayG. Moorna Formation
<200yr
<6400yr6400-1 8,000+ yr
?Mid. Pleistocene
4000-6000 yr
Pliocene/PleistoceneLate Pliocene
? Moorna Formation
The Fluorine Indices as a group are high,
but the reason for this is not understood. How-ever, they show an increase with age, and the
percentage of nitrogen falls off with age. There
is thus a general relative dating, but the chief
value of such tests is to discover samples that
are out of context. Thus clearly sample 8 is
wrong. The bone belongs to a bettong, a bur-
rowing marsupial, so perhaps it is intrusive. Thehuman remains in the Rufus Formation
(samples 11 and 12) are also clearly intrusive.
The disconformity between the Blanchetown
Clay and the formations in the incised river
valley is made clear by the jump in fluorine in-
dices and the drop in nitrogen percentages. TheBlanchetown Clay and the Moorna Formation
are conformable, the former overlying the latter.
This is reflected in the very similar figures.
Sedimentation
"To see a world in a grain of sand" William
Blake (Auguries of Innocence)
The development of geophysics made pos-
sible the discovery of the fundamental structure
of the earth's crust. Investigations led to the
Fluorine
.
% SampleIndex Nitrogen number2.28 0.226 10
6.86 0.026 96.62 0.043 6
9.73 0.020 7
1.89 — 8
5.53 0.045 45.09 0.041 5
2.45 0.041 11
1.81 0.118 12
16.94 0.019 1
16.42 0.013 220.58 0.007 3
14.09 — 1417.15 — 15
19.06 — 1614.80 — 13
differentiation of the heavier oceanic crust and
the lighter continental crust. So evolved the
concept of the continents as rafts of sial floating
in a sea of sima.
Later, the demonstration of a far greater
horizontal mobility for the crust than pre-
viously envisaged, the discovery of mid-oceanic
ridges with evidence of spreading, and the need
for some other explanation of tectonics and
vulcanism than as expressions of a shrinking
earth, led to new credence being given to
Wegener's hunch of continental drift. Wegener
did not prove it, nor is it now proved. How-ever, it is true that the supporting evidence has
been very greatly increased. During the past
10 years there has been such a rapid accumula-
tion of data and so fast a flurry of changing
ideas that this is clearly no time for dogmatism.
The calving of land masses (e.g. Australia
from a southern supercontinent), and the col-
lision of continents are vast traumatic events
which must leave major lesions on the skin of
the earth. On the other hand, there are manylong slow warpings of the crust with associated
accumulation of sediments that are of quite a
different order of dynamics. Such is the Murray
62 EDMUND D. GILL
Basin. Although cheek by jowl with the massive
fault system that yields the Mt. Lofty Ranges of
Precambrian sediments, the Murray Basin has
very slowly subsided over more than 120 m.y.
The mean rate through this period in the deep-
est part of the basin (and so the fastest sinking)
is only 1 m per 120,000 yr. At this maximalrate, the land would sink only 8*3 cm in the
whole of the Holocene (10,000 yr). It is nowonder that the surface of the Murray Basin is
so flat and its dynamics so low. This is the
background against which to view the grain
size analyses. However, the factor of tectonicity
(cf. Krumbein and Sloss 1963) has been quali-
fied by climatic change (see section on palaeo-
climatology), for a humid region has becomea semi-arid one.
Grain Size Analyses
Unless otherwise stated, the sieving has beencarried out by Mr. K. G. Simpson and assistants
under the direction of Dr. A. W. Beasley,
Curator of Minerals, National Museum of Vic-
toria, whose research lies in this field. Theintegration and interpretation of these results
are the writer's responsibility. I am indebted to
Dr. Beasley for his considerable assistance.
Mrs. J. Dudley checked the calculations anddrew most of the histograms.
Enquiries indicate that communication wouldbest be served for those in the many disciplines
interested in this report if the results are givenin the form of histograms. These are numbered,and in general follow from the oldest to the
youngest sediments. Wentworth's size classifica-
tion is used.
Parilla Sand
This formation outcrops in the Murray Rivercliff beside the Kulcurna Station homestead(Fig. 19). The sediments are gray (unoxidized)and without reaction to acid. The formationwas sampled at two levels:
1. Clayey sand at base of cliffs. SampleCHG/78.
2. Same 3 6m from top of formation (6 mexposed). CHG SS 115.
Histograms 1-2 (Fig. 49) show well-sorted
sediments with better sorting higher in the
I -5 -26 12S 06J m 4 z i 5 :s -i2» -ot)
Fig. 49—Grain size analysis histogramsof Parilla Sand. 1 = Grayclayey sand at base of Mur-ray R. cliffs, Kulcurna Station
homestead, Cal Lai. N.S.W.Sample CHG 78. Site 2 =Same site. 3-7 m below topof formation. Sample CHGSS115. For section see Fig. 19.
formation. Histogram 2 is remarkably like that
for the overlying Chowilla Sand (Histogram 9,
Fig. 53) which suggests that one was derived
from the other, or both came from a commonsource by a transport system of like dynamics.
Moorna Formation
Histograms 3-6 (Fig. 50) are of MoornaFormation sediments in the section at the W.end of Fishermans Cliff (fossil bone site).
Histograms 4-5 are of the same sample, andtheir close similarity is taken to mean reliability
of analysis. They are unusually coarse for this
region, and represent a riverine channel deposit.
The deposit cuts out laterally. The base of the
Moorna Formation outcrop at this site is finer
(Histogram 3) with the medium and fine sandcolumns dominant as against the dominantcoarse sand column for the sediments above.
Histogram 6 represents an unusual sedimentbecause many of the grains are coated withdolomite (see Segnit, this volume). The in-
ference is that these were eroded from a bedof sandy dolomite such as occurs at TripleSwamp Gulch (Fig. 24). The grains are wellrounded. Some fines make the histogram slightly
bimodal. Figure 51 shows the change in grain
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 63
5 25 -125 '063 1 -5 "25 125 063
Fig. 50—Grain size analysis histograms of Moorna Formation sediments (see
also Figs. 50-51). 3 = W. end of Fisherman's Cliff, N. bank. MurrayR., Moorna Stn., W. of Wentworth, N.S.W. Fig. 30. Sample CHGSS113. Site 13. 4 = Same site, clean sand at top with some grains
dolomite-coated. See Fig. 51. The level is between the lower clayey
part of the Moorna Formation and the overlying yellow Chowilla
Sand with fossil tortoise. The sample is part of the matrix that
carried the fossil bones described by Marshall in this Memoir, it
came from c. 105 cm above the clayey sand (CHGSS 111).
size analysis when the dolomite is stripped by
acid treatment.
Histograms 7-8 (Fig. 52) are of Moorna
Formation sediments in two places other than
Fisherman's Cliff. A brown (7 5 YR 5/6) sand
at the base of the exposed section at Triple
Swamp Gulch W. of Fisherman's Cliff gave
histogram 7. The dominant grade is medium
sand. Histogram 8 is of the thick light-gray
sands in the Merbein Cliffs (Fig. 36). The
grades of sand vary. This sample is from 06 mbelow the siliceous paleosol in the section at
The Lookout. The spread of grades is rather
similar to that of the Chowilla Sand into which
it merges. The coarser sediments with current
bedding occur lower in the section.
Chowilla Sand
Histograms 9-16 (Fig. 53) cover this forma-
tion, following sites from W. to E. No. 9 is
remarkably similar to that of the underlying
Parilla Sand (No. 2) and suggests derivation
from it, or the same source. No. 10 shows the
variation in the top of the formation. The
major part of this deposit is homogenous with
a slightly columnar structure. The top has sub-
Gram
16 30 60 120 240 Pan
BSS Sieves
^B Before acid treatment
l^'VI After acid treatment
Pig, 5i—Grain size analysis histo-
gram of sediments fromthe Moorna Formation at
the W. end of Fisherman's
Cliff, Moorna Station,
N.S.W. Site 13. Histo-
grams 4-5 show the analy-
sis of natural sand, while
this figure shows the effect
of stripping off the dolo-
mite coatings.
64 EDMUND D. GILL
Fig. 52—Histograms 7-8 of grain size
enalysis of Moorna Forma-tion sediments. 7 = Whitesand 0-6 m below top offormation (i.e. below the
paleosol) in the cliff section
at The Lookout, Merbein,V. Fig. 36. Sample CHGSS108.
parallel bedding. No. 11 is from NampooStation in the Murray River cliffs where the
Chowilla Sand is a narrow bed in the Blanche-
town Clay sequence.
Dr. A. W. Beasley reports on the mineral
composition of samples 13 and 14 as follows:
"The sand samples are composed essentially of
light mineral grains, most of which are quartz.
Feldspar grains are generally cloudy from al-
teration, and some are almost opaque; they are
often lath-shaped. White mica occurs in minor
amounts; it is more common in the finer size-
grades of the samples. The quartz grains in
samples 13 and 14 are commonly iron stained
in a patchy fashion, and inclusions are very
numerous. The larger grains are subrounded to
rounded, while the finer size particles are usual-
ly subangular. The degree of roundness of most
grains indicates that they have had a fairly long
detrital history."
Heavy Minerals
Dr. Beasley reports that "the assemblages
consist essentially of the same species, but
there are differences in their relevant propor-
tions, the significance of which could only be
determined by more numerous analyses. Thesuites indicate granitic and metamorphic
source rocks. The opaque minerals are relatively
common, while the non-opaque heavy minerals
include zircon, tourmaline, hornblende, apatite,
rutile, sphene, anatase, staurolite, garnet, epi-
dote, andalusite, cassiterite and biotite. Hiedegree of roundness of the heavy mineral grains
varies considerably; most are subrounded or
subangular. The weight percentages of heavy
minerals in the 25 mm to 0T25 mm size
grains are as follows:
Sample 13 020%Sample 14 0-25%Sample 17 013%."
Sample 14 has the lowest degree of sorting
of all the Chowilla Sand samples tested. This
may be due to being so close to the clay facies.
At Bone Gulch there is a series of alternating
thin beds of clay and sand. The sands have con-
centrations of fossil bones, and the highest
percentage of heavy minerals. Ostracods are
common, indicating shallow water. Some of the
bones are worn. It is considered likely that
these beds were deposited by an oscillating level
of the lake, the higher levels washing the sands
and concentrating bones and heavy minerals.
The partly articulated bones of a large dipro-
todontid were found in the base of the Blanche-town Clay. The animal may have been mired in
shallow water.
Histogram 15 is of a sample from the Yeltacliffs which likewise has a comparatively poordegree of sorting . Histogram 16 is of a samplefrom the Merbein cliffs, and shows a dominanceof medium and fine sand grades seen also in
other samplings of this formation.
Blanchetown Clay
Histograms 17 and 18 (Fig. 54) are of sam-ples from this formation (with the clay re-
moved). Sample CHG SS/42 (No. 17) is fromjust above the ossiferous Chowilla Sand at BoneGulch, Moorna Station, N.S.W. The sediments
are poorly sorted with three modes—clay
(69%), silt and fine sand. Histogram 17 for
material greater than clay size is bimodal, witha primary mode in the fine sand grade and alesser one in the silt. Dr. Beasley reports "The
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 65
5 -25 125 063
Fig. 53—Histograms 9-16 of grain size analysis of Chowilla Sand. 9 = Upper part of Murray R. cliff
at Kulcurna Stn. homestead, Cal Lai. N.S.W. Site 3. Fig. 19. Sample CHG 38. 10 — Samesite, but sample from cliff top. CHG 77. 11 = Under Blanchetown Clay at foot of cliffs
below Nampoo Stn. homestead. E. of Cal Lai, N.S.W. 12 — Sand below Blanchetown Clay,Middle Gulch. Noola Erosion Amphitheatre, Noola Stn., NW. side of Lake Victoria. Site 6.
W. end, Fisherman's Cliff, Moorna Stn., N.S.W. Sample CHGSS 43, 15 = Middle of sandformation at top of cliffs, S. bank Murray R., at Yelta, Vict. See Gill on Palaeopedology, thisMemoir, Fig. 1. 16 = Chowilla Sand between Blanchetown Clay and Moorna Formationin Merbein Cliffs, Vict., at The Lookout. Fig. 36. Sample CHGGS 114.
material greater than clay size is mainly quartz.
Although inclusions are numerous in the quartz,
the grains are comparatively free from iron
staining. They are cleaner than in the sand
samples. Ostracods are fairly common. Felspar
grains are cloudy from alteration. Flakes of
white mica are present but not common. Thesuites of heavy minerals indicate granitic and
metamorphic source rocks. The weight per-
centage of heavy minerals in the . 25 mm to
0063 mm grade is 0-13. The underlying
Chowilla Sand has 0-25%".
A sample of Blanchetown Clay from the
Murray River cliffs at Nampoo Station, E. of
Cal Lai, N.S.W. had 90% clay, so the range of
clay content in this formation is considerable.
Within the Blanchetown Clay are bands of sandwhich vary from white (well-washed) to green-
ish gray (containing clay).
Histogram 18 (Fig. 54) shows the sampleCHG 151a (after washing) from a sandy layer
in Noola erosion amphitheatre on the NW.shore of Lake Victoria. The analysis is domin-ated by the fine sand grade, and so represents
66 EDMUND D. GILL
A 2 1 & » 1M <m 4 2 1 6 26 12b OCJ
Fig. 54—Histograms 17-18 of grain size analy-
sis of washed samples of Blanche-town Clay. 17 = Clay from BoneGulch. Moorna Stn. (Fig. 34), W. of
Wentworth, N.S.W. 18 = Sandy layer
in this formation with CorbicuIina y
Noola Erosion Amphitheatre, NoolaStn.. NW. side Lake Victoria (Fig.
17), N.S.W. Sample CHG 151a.
an environment of low dynamics. The bed hadnumerous shells of the bivalve Corbiculina.
Woorinen Formation
1. Berribee Station Dunes. An auger hole
put down through an E-W. dune provided the
samples which yielded the histograms 19-26
(Fig. 55). They all show good sorting with
fine sand being the dominant grade. All the
histograms are unimodal except 24 and 26which are bimodal. It would be interesting to
know if the bimodal zones can be traced from
dune to dune.
2. Hattah Dune. Histogram 27 (Fig. 55) of
sediment from the Hattah district, but in the
same system of E-W. dunes, gave a similar re-
sult.
Histogram 24 (sample 36 from 3 4-4 m) is
the best sorted (SO=13) yet is bimodal with
21-59% silt. The base of the dune is repre-
sented by histogram 25 (sample 37 from 4-6-
5-5 m) which is skewed to the fine side.
Samples 36 and 37 had 119% and 570%of clay removed before the grain size analyses
were commenced. The clay content is believed
to be ( 1 ) partly blown in with the original sedi-
ments as clay skins on sand grains, and as ag-
gregations, and (2) partly added later from
dust storms. The increase in clay content at
the base is believed to be due (1) partly to
washing down of clay by rain water, and (2)
partly to the dune overlying a very clayey sub-
strate which probably contributed a good deal
of clay in the initial stage of dune building.
Histogram 26 (sample 38 from 5-5-5 7 m)represents part of the old terrain on which the
dune was built. It shows the typical dominant
fine sand column, but with a silt column almost
as high. The histogram is thus bimodal. To gain
the sample for grain size analysis 57*92% of
clay was removed from the field sample. Afloodplain lagoon or a lake is indicated, be-
cause 83% of the field sample was fine enoughto pass through BSS sieve 240.
Rujus Formation
1. Dunedin Park Station, N.S.W. Figure 37
provides a surveyed section across the margin
of an outlier (number 1 in Fig. 5 ) of this forma-
tion near Frenchman Creek. This outlier is a
remnant of a higher Pleistocene terrace of the
Murray River. Its sediments are now oxidized
to a red colour, that contrasts with the greenish
gray sediments of the present floodplain. Nodoubt the Rufus Formation was once that colour
too. Four auger holes were sunk along the sec-
tion line. For logs see under Rufus Formationin stratigraphy section. Histograms of the sedi-
ment analyses are given in Fig. 56 as follows:
Histogram 28 Outlier 1, Auger 1, 0-20cm29 20-46cm30 61-69cm31 0-94-1 -02m32 1-22-1 -52m33 1-80-1 -88m34 1-88-2- 13m35 2-21-2-36m36 2 -67-2 -74m37 2 -74-2 -90m38 30O-3-O7m
Histogram 39 Outlier 1, Auger 2, 0-1 5cm40 53-61cm41 86-94cm42 1-22-1 -35m43 1-75-1 -83m44 3 -76-3 -84m
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 67
mm 4 2 1 -5 -26 125 '063 2 1 -5 "25 125 063 mm 4 2 1 '5 "25 125 063 4 2 1 '6 '25 '125 063
Fig. 55—Histograms 19-27 of grain size analysis of sediments from the Woorinen Formation
(Pleistocene E-W. dune system). 19-26 from Auger hole 2 through dune c 3 km SSW.of Berribee Stn. homestead, Vict. CHGSS sample nos. in brackets.
19 = — 0-38 m (32), 20 = 6 — 0-9 m (33), 21 « 2-3 — 2-6 m (34), 22 =2-6— 3-2 m (35), 23 = Repeat of 22, 24 = 3-4 — 4 m (36), 25 - 4-6 — 5-5 m(37),26 = 5-5 — 5-7 m (38), 27 = Sediment from 2-6 — 2-9 m (118) in auger hole through
E-W. dune, Calder Highway between mileposts 309 and 310, near Hattah, N. Vict.
end of Brown's Paddock (Site 17) Keera
Station, Victoria, that at first glance look like
dunes. They are extensions from a Pleistocene
terrace remnant. Three auger holes were sunk
as follows
:
Auger 1 46+cm of gray well-washed mobile sand(which the auger could not lift) at the N.end of section (Fig. 38), local channel.
Auger 2 For site see Fig. 38.
50cm Red (2-5 YR 4/8) slightly clayey sandforming hardpan (surface sand deflated)
50cm Reddish brown (5 YR 5/4) and such clayeysand with glaebules of earthy carbonate up to
2 cm diameter. Gradually becomes grayer
2. Keera Station, Victoria. Figure 38 is a cross- gS-KriSXcR^JL^losection of Rufus Formation structures at the E. 90cm
Histogram 45 Outlier 1,
46
Auger 3, 3-07-3-15m3-35-3-51m
Histogram 47 Outlier 1,
48
Auger 4, 3-51-3-63m4-ll-4-27m
Histogram 49 Outlier 1,
5051
52535455
Auger 4, 0-1 5cm36-46cm
0-99-1 -07m1-30-1 -37m1-57-1 -65m1-83-1 -88m2 -29-2 -36m
Histogram 56 Outlier 2,
57
58
Auger 1. 53-6 1cm1-32-1 -42m2 -44-2 -54m
68 EDMUND D. GILL
pig. 56—Histograms 28-82 of sediments from the Pleistocene Rufus Formation (riverine). CHGGSsample numbers in brackets. Samples from auger holes in residual marked 1 on mapin Fig. 5, Dunedin Park Station, N.S.W.28 = — 0-2 m (74), 29 = 2 — 46 m (75), 30 = 0-6 — 0-68 m (76), 31 =0-94 — 12 m (77), 32 = 1-2 — 1-5 m (78), 33 = 1-8 — 19 m (79), 34 = 1-9— 2-1 m (80), 35 = 2-2 — 2-4 m (81), 36 = 2-67 — 2-74 m (82), 37 = 2-74— 2-89 (83), 38 = 3 — 3 07 m (84). See histograms 57-58.
Auser hole 239 = — 0-15 m (85), 40 = 0-53 — 6 (86), 41 = 0-86 — 0-94 m (87),
42 = 1-22 — 1 35 m (88), 43 = 1-75 — 1 83 m (89), 44 = 3-76 — 3-83 m (91).
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 69
nun 4 2 1 '5 -25 125 063 4 2 1-5 '25 -125 0634 2 1 -9 -25 -125 -OSa mm 4 2 1 -E -25*125-063. mm 4 2 1-5 25 125-063
Auger hole 3A45 = 3.5 _ 3-63 m (100), 46 = 4-1 — 4-3 m (101).47 = — 0-15 m (92), 48 = 0-36 — 46 (93), 49 = 0-99 — 1 07 m (94),50 = 1-3 — 1-37 m (95), 51 = 1-57 — 1-65 m (96), 52 = 1 83 — 1 88 m (97),Auger hole 3
53 = 2-29 — 2-36 m (98).
Just W. of residual marked 2 in Fig. 554 - Q.53 _ 0-6 m (102), 55 = 1-32 — 1-42 m (104), 56 = 2 44 — 2-54 m (105).Auger 1, residual 1 (see histograms 28-38)
57 = 3-04 — 314 m (106), 58 = 3-35 — 3 5 m (107).
70 EDMUND D. GILL
4 2 1-5 25 125 093 mm 4 215 -25 125 063 mm 4 2 1 "6 25 -125 063 mm 4 21-5 15 -125 063 mm 4 2 1 8 M "125 063
4 2 1-5 -25 125 063 mm 4 2 1 5 -25-125-063 4 2 1 "5 25 125 063 mm 4 2 1 S 25 125 063
Surface soil, residual marked 1 in Fig. 5
59 = Loose red sand at surface, 60 = Sardpan below.Brown's Paddock (Site 17), Keera Stn., Vict. (Fig. 38), auger hole 1
61 ~ 0-76 — 0-91 m, 62 = 1-98 — 2-13 m, 63 = 2-28 — 2-44 m.Auger hole 464 = — 0-41 m, 65 = 0-51 —- 0-76 m, 66 = 1-02 — 1-32 m, 67 =78 = 2-44 — 2-69 m.Lybra Paddock (Site 16), Keera Stn., Vict. Auger hole 1
69 - — 0-43 m, 70 - 0-43 — 0-91 m, 71 - 0-91 — 1-6 m, 72 =73 ~ 2-74 — 2-92 m.
1-83 — 206 m,
1-6 — 1-68 m,
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 71
4 2 15 -26 125 063 nun 4 2 1 -9 '25 125 053 mm 4 2 I -5 25 125 063
mm 4 2 1-5 -25 -125 063 mm 4 2 1 -6 -25 125 063
Auger hole 274 — o-l — 0-46 m, 75 = 0-46 — 0-66 m, 76 = 0-66 — 117 m (samples 74-76 werefrom the middle of the depth ranges given), 77 = 1-17 — 1-27 m, 78 — 1-27 — 1-55 m79 s= 2-29 — 2-54 m, 80 — 2-74 — 3-5 m, 81 = 3-5 — 3-66 m, 82 — 4-57 — 4-78 m'
112m Zone of incipient mottles, some areas tend-ing towards grayish green and others to-
wards brown. At 2 m from surface gradualchange to less clayey sediment
51cm Light yellowish brown (2-5 Y 6/4) sand.Probably more oxidized because more per-
meable13cm More clayey and with light gray (10 YR
7/2) mottles.
84+cm Light yellowish brown (2-5 Y 6/4) sand.Total depth 3 -6m.
Auger 3 33 -2m N. of skeletons (C14 site).
30cm Red (10 YR 5/6) loose sand without car-
bonate45cm Red hardpan of same sand with clay; no
carbonate28cm Light gray to light brownish (10 YR 6/1)
clay with off-white masses of earthy car-
bonate up to l-5cm in diameter. Selenite
crystals (needles)
94cm Similar sediment almost without carbonate
46*cm Same with pyrolusite.2 -43m Total depth.
Auger 4 On section line (Fig. 38) near samphire flat.
25cm Reddish brown (5 YR 5/3) slightly clayeysand forming a hardpan, merging toSame with carbonate merging toPink (5 YR 7/3) to pinkish gray (7/2)clayey sand with glaebules of earthy car-bonate pinkish white (7-5 YR 8/2), merg-ing to
Light brownish gray (10 YR 6/2) sand withless carbonate merging toClayey sand, some colour, reduced carbonate;pyrolusite and gypsum appearing; mergingto
Pale red (2-5 Y 6/2 and such) clayey sandLight brownish gray (2-5 Y 6/2) to light
7/2) sand with yellow (2-5 Y 7/3 at 2-8mfrom surface.
Total depth 2- 31m.
AugerS On section line (Fig. 38) in middle ofsamphire flat.
33cm Red sand blown and/or washed over flat,
and some gray sand intermixed
5cm30cm
15cm
30cm
107m19 +cm
72 EDMUND D. GILL
15cm Yellowish red (5 YR 5/6) clayey sand, somepyrolusite at top.
43cm Pinkish gray (7-5 YR 3/2) sandy clay withnumerous patches of earthy carbonate
56+cm Light brownish gray ( 10 YR 6/2) sandy claywith a little carbonate and with selenite
(rich at 107m from surface)l-47m Total depth
Auger 6 On crest of south ridge of section line (Fig.
38).13cm Light red (2-5 YR 6/6) slightly clayey sand
forming hardpan20cm Red (2-5 YR 4/6) sandl-27m Pink (7-5 YR 7/4) sand with earthy car-
bonate glacbules up to 4cm diameter.Amount of carbonate reduced from 84 cmfrom surface merging to
46cm Reddish yellow (7-5 YR 6/8) sand withfinely divided carbonate
18cm Dark brown (7-5 YR 4/4) sand withselenite and pyrolusite (which makes thesediment darker), merging over 5cm to
8cm Pink (7-5 YR 8/4) sand with no appreciableamount of carbonate
119m Light brown (7-5 YR 6/4) clayey sand
—
a marked increase in clay content25 +cm Reddish yellow (7-5 YR 6/6) sand.3- 76m Total depth.
The sediments from anger holes (2) and(3) were subjected to grain size analysis andthe results are given in histograms 59-68 (Fig.
56) as follows:
Histogram 59 Site 1, loose red surface sand60 Site 1, auger 3,
61
62
63
64 Site 1, auger 2,
65
6667
68
0-38cm76-9 lcm
1-98-2 -13m2 -29-2 -44m
0-4 lcm51 -76cm
1 -02-1 -32m1-83-2 06m2-44-2- 69m
The auger holes prove that the profile is a
stable one except for the mobile surface sandwhich was probably stable also before occupa-tion by flocks of sheep. The profile has beenstable long enough for the carbonate soil to
develop, which (judging by the soil on the E-W.dunes) is at least 16,000 yr old. This soil occurs
(with modification) on the flat as well as the
ridges (Fig. 38), so no strong erosion can haveoccurred there during that period. Local in-
formation is that floods reach the end of the flat
but do not penetrate between the ridges across
the section line. The erosion of the formerterrace (judging by the soil) must therefore
have occurred prior to 16,000 yr ago. The
terrace was eroded before the Lake Victoria
Sands were deposited because they onlap out-
liers. However, more than halfway up in these
largely subparallel blowout sands is a paleosol
with a midden dating 17,530 yr B.P. If that
figure is doubled, a chronology like that of the
Lake Mungo lunette is obtained. So the erosion
of this system may have occurred in the earlier
half of the Last Glacial. It would take a long
time because hundreds of square kilometres of
this terrace have been excavated, as Fig. 4shows.
The sediments are all fine and very well
sorted, but the stratigraphy is very variable as
one would expect in a riverine environment.This applies to the Coonambidgal Formation,and to the sediments being laid down at the
present time. Figure 61 shows variation in clay
content from 50% to 68*23%, and in car-
bonate content from to over 30% by weight.
The grain size analyses were done on clay-freeand carbonate-free sediments.
On Keera Station, auger holes were sunk alsoat Site 6 (Lybra Paddock) where manyAboriginal skeletons were excavated, and somedated by CI 4. They came from burials intrusivein the surface of the eroded Rufus Formation.Figure 57 shows auger 1 at the level where theskeletons were found, and auger 2 at the level
SUL \(
KU IS FORMATION
-
hl ,i,rtrerta deposits
1' fwuind Saiul and
Auger hole
2
<1au-\ Sand
Fig. 57—Eroded Rufus Formation in LybraPaddock, Keera Station, beside Wall-polla Creek, Vict. The site of Angerhole 2 is the 1956 floodplain of theMurray River. All sediments oxidizedexcept 117 m of sediments on theabove floodplain. Both auger holeswere 4-8 m deep; only Auger hole 1
is shown fully. Line between augerholes = 57° magnetic.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 73
of high floods as in 1956. This level was sur-
veyed to a bench mark on the River Road bySir Robert Blackwood and Mr. Gerald DouglasThis showed it to be about the same height as
the site of auger 1 in Brown's Paddock (Fig.
38). The logs of the auger holes are as follows:
Auger 1 On terrace.
43cm Strong brown (7-5 YR 5/6) micaceous (asall the sands here) grading to
48cm Reddish yellow (7-5 YR 6/6) sand withpatches of earthy carbonate up to 0-5cmdiameter, grading to
1 -55m Similar sediments with carbonate concretions(irregular, elongate)
l*47*m Similar sediments with only traces of car-bonate.
4 -77m Total depth.
Auger 2 On floodplain 0-82m lower.10cm Light gray (10 YR 7/1) silty sand36cm Dark brown (7-5 YR 3/2) to very dark
gray (3/1) clayey sand, grading to20cm Grayish brown (2 5 Y 5/2) clayey sand
with patch of earthy carbonate hard enoughto grate an auger at times, grading slowly to
51cm Yellowish brown (10 YR 5/6) slightly
clayey sand, carbonate fading out at 84cmfrom surface, sharp break to
10cm Very pale brown (10 YR 8/3) washed sand,probablv a channel, sharp break to
28cm Light gray (2-5 Y 7/3) to pale yellow (7/4)slightly clayey sand, sharp break to
10cm Very pale brown (10 YR 7/3 washed sand,probably a channel, rapid change to
53cm Pale yellow (2-5 Y 7/4) slightly clayeysand
10cm Pale brown (10 YR 7/4) washed sand(channel?)
25cm Light yellowish brown (10 YR 6/5-6) tobrownish yellow sand
20cm Mottled greenish gray (5 GY 6/1) to yellow-ish brown sand with clayey accretions, grad-ing to
97cm Yellowish brown ( 10 YR 5/4) washed sand46cm Darkish reddish brown (5 YR 2/2 and such)
washed sand, sharp break to15cm Light gray to greenish gray (5 Y 7/1 to
5 GY 7/1) washed sand with yellowish red(5 YR 5/8) mottles, sharp break to
4 Tern Yellowish red (top 3cm) to yellowish brown(10 YR 5/4) washed sand.
4- 70m Total depth.
Grain size analyses (Fig. 56) were carried out
on sediments from these two drillings as follows:
Histogram 69 Site 16, auger 1, 0-43 cm70 43-91 cm71 0-91-l-60m72 l-60^1-68m73 2-74-2-92 m
Histogram 74 Site 16, auger 2, 10^-46 cm75 46-66 cm76 0-66-M7m77 1-17-1 -27 m
7879808182
I -27-1 -55 m2-29-2-54m2-74-3-51
m
3*51-3-66
m
4-57-4-77
m
Monoman Formation
The solitary sample (CHG 55) from this
formation consists of the matrix round the
fossil log used for radiocarbon dating. It camefrom the trench sunk for an experiment on abitumen curtain designed to prevent underflowat the proposed Chowilla Dam (Figs. 39-40).
Histogram 28 (Fig. 56) shows a channel de-
posit with the dominant column being very
coarse sand (1-2 mm).
Coonambidgal Formation
This formation includes tine-grained lagoon(or oxbow) and floodplain sediments, channelsands and channel border dunes. Histogram 29(Fig. 56) is of alluvial on the SW. shore of
Lake Wallawalla that formed the matrix of
skull 61 -A. It is an admixture of all gradesfrom coarse sand to silt inclusive. There is adoubt whether these sediments should be in-
cluded in the Coonabidgal Formation, but timeto investigate the point has not been available.
By contrast, histogram 30 (Fig. 57) is ofalluvium at the base of auger hole 1 sunkthrough a channel border dune. It is essentially
a silt, being skewed strongly to the fine side.
Histogram 31 is of the matrix of skeleton 53-Aat the W. end of Lindsay Island, Berribee Sta-
tion, NW. Victoria. Once again there is a typical
dominance of the fine sand grade, yielding astrongly unimodal histogram.
Figure 12 shows a section of a channel bor-der dune in the Coonambidgal Formation onLindsay Island, Berribee Station, NW. Vic-toria. The logs of auger holes 1-3 are given in
the geomorphology section where dunes are
discussed. Auger hole 4 showed the following:
10cm Pinkish gray (7-5 YR 6/2) well sortedsand washed from dune (surface).
36cm Light gray (10 YR 7/1) poorly sorted sandwith grain sizes coarser than in dune.
33cm Lighter gray (N7) sand, very poorly sortedand with clay balls up to 7cm diameter.Gradual change to
8cm Light brownish gray (2-5 Y 6/2) sand witha few streaks of light olive brown (2-5 Y5/4).
74 EDMUND D. GILL
28cm Off-white (near N8) well-washed sand, verydifficult to bring up.
61 +cm Same but light gray (2-5 Y 7/2).
An analysis of the N8 sand from this riverine
sequence was carried out (histogram 87). Theusual dominant column appears, but offset
one grade to the coarse side because it is froma channel deposit. The mobile sand preventedfurther boring. Berribee auger hole 5 was sunkE. of the section shown in Fig. 11. The site
is a red gum hollow, which is filled by present
flood waters situated about 110 m SSE. of the
peg at auger 1.
10cm Dark brown (10 YR 4/3) well sorted sand.Juvenile soil formed by accumulation ofhumus, merging to
10cm Grayish brown (10 YR 5/2) similar sand,merging to
10cm Pale brown (10 YR 6/3) sand, merging to28cm Light gray (10 YR 7/2) well washed sand,
merging to63cm Off-white (2-5 Y 8/2) mobile sand gradu-
ally changing to23cm Very pale brown (10 YR 7/3) mobile sand,
gradually changing tol-22m Light yellow brown (10 YR 6/4) mobile
sand15cm Pale brown (10 YR 6/3) sand, and light-
gray (5 Y 7/1 to 5 GY 7/1) clayey finesand merging to
36cm Very pale brown (10 YR 7/3) sand13cm Same with gray clayey fine sand as above8cm Pale brown (10 YR 6/3) sand without gley8cm Same with patches of yellow (10 YR 7/6)13cm Brown (10 YR 5/3) clayey sand25cm Same with gray clayey fine sand as above.36cm Brown sand without gley,33 +cm Brown with mottles of gray and strong
brown (7-5 YR 5/6) concentrating in placesto dark reddish brown (5 YR 3/2) sand,but the latter disappears 15cm from thebottom, leaving brown and gray sand only.
4.53m Total depth.
An analysis was carried out of a sample 71-74cm from the surface (off-white mobile sand),and the result appears in histogram 88 whichis bimodal. Dominance moved back to thefine sand grade. Sand is blown from the chan-nels to form the dunes, and sand from the dunesis washed back into the channels. The gradeof sediments dominating the channel depositsis limited by its dynamics. The sands are wellsorted (So — 1-46) as usual.
Nulla Nulla Sand
This is the lower member of the Lake Vic-toria Sand (lunette). Fig. 58 presents three
histograms characterized by a dominant columnof fine sand—the typical grade in this river sys-
tem of low dynamics. The sediments are finer
than are usually found in dunes, but this is whatthe river transport provided. The light weight of
the sand grains, the limited supply brought in
by the river, and the persistent seasonal W.winds are no doubt the chief contributory fac-
tors to the subhorizontal position of most ofthese sands. Ample winds moving a light sandof limited quantity kept the dune in a state ofalmost constant blowout.
Histogram 89 (Fig. 58) is from TalgarryStation just S. of the border fence with NullaStation, below our bench mark 1 (CHG SS110). Histogram 90 is from a clayey horizon(CHG 98) with badlands erosion on NullaStation (N14/15). Histogram 91 (Fig. 58) is
for the matrix of a fossil marsupial from Tal-garry Station, and is dominated by the veryfine sand grade.
Talgarry Sand
This is the upper member of the Lake Vic-toria Sand (lunette). Histograms 92 and 93(Fig. 61) show two types of sediment foundin the Talgarry Sand that have parallels in theunderlying Nulla Nulla Sand.
Sediments in relation to Facies. The sorting co-efficients of the sand fractions relative to facies
83
5 26 126 M3
Fig. 58- -Histogram of grain size analysis of Mono-man Formation sediments from the cut-offtrench at the Chowilla Dam location (Site1). Sample CHG 55.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 75
mm 4 2 1-5 25 725 -063
4 2 1 *6 '28 126 -063 a j l izs QC3
Fig. 59—Histograms 84-88 of grain size analyses of Coonambidgal Formation sediments.84 = Matrix of Aboriginal skeleton 61A from SW. side of Lake Wallawalla (Figs. 14-15).Sample CHGSS 118. This is doubtfully referred to this Formation. 85 = Mottled clayfrom bottom of auger hole 1, c. 3 km SSW. of Berribee Stn. homestead under WoorinenFormation dune. This is almost certainly not Coonambidgal but the formation to whichit belongs has not yet been determined. Sample CHGSS 31. 86 = Matrix of Skeleton53A, W. end of Lindsay Island (Site 19). Coonambidgal channel border dune. SampleCHA 95. 87 = Off-white (2-5Y N/8) sand from auger hole 4, Lindsay Ts., Berribee Stn.N.S.W. (Fig. 12) 0-86 — 114 m. Sample CHGSS 39, 88 = Same site, auger hole 5 sunkon E. side of channel border dune (Fig. 12) 0-66 — 0-74 m. Sample CHGSS 40.
are as follows (the number of So determina-tions is given in brackets):
1. Riverine floodplain
3.
1 3-1 171-20-2-55
Parilla Sand (2)Rufus Formation (52)
(omitting So 4-3)
Coonambidgal Formation (5 ) 1 22-1 46(omitting So 2 10 which doubtfully be-
longs to the formation)
Riverine channel
Moorna Formation (5) 1*25-1-54
Chowilla Sand (6) 114-1 68Blanchetown Clay, sandy facies (1) 1-27
Lacustrine
Blanchetown Clay (1)
DuneWoorinen Formation (9)Nulla Nulla Sand (3)Talgarry Sand (2)
1 29-1-521-53-1-60
1 44-1-77
The whole series is remarkably fine which can
be attributed to long transport and no fresh
sediments brought in during the long flows
across dry country. The degree of sorting is
remarkable. This can be attributed to the lowdynamics cutting out the larger fractions, andthe long transport ever improving the sorting.
There is an enormous range in the percentagesof carbonate and clay. Accumulation of car-bonate is largely a pedogenic function. Thevariations in the clay content of riverine alluvia
are such as are to be expected. In this semi-aridcountry, windblown fine sediments have beenadded later to waterlaid sediments.
Denudation. The mean denudation rate is re-markably low. Langbein and Schumm (1958)have studied the quantitative relationship be-tween sediment load and annual precipitation,
concluding that maximal sediment yield occursin areas of 25-30 cm rainfall, to which ourstudy region belongs. The authors measuredsolid load only. Ahnert (1970) points out thatseasonal distribution of rainfall is important,and it certainly is in our Project area. Ahnertdeals also with the significant factor of relief.
The mean denudation rate in mid-latitude riverbasins is directly proportional to mean basinrelief. The very flat Murray Basin has minimaldenudation. Many relict features such as the
76 EDMUND D. GILL
Fig. 60—Histograms 89-91 of grain size analyses of sedi-
ments from the Nulla Nulla Sand member of the
Lake Victoria Sand Formation (Figs. 42-45).
89 = Nulla Nulla Sand from below bench mark1 on S. (Talgarry Stn.) side of the boundaryfence (Fig, 16). Sample CIIGSS 110, 90 Clayeysand with herringbone erosion. Gulch N14 (Fig.
16) interpreted as a fossil claypan. Sample CHG98, 91 = Matrix of marsupial fossil from NullaNulla Sand, Lake Victoria Lunette, Talgarry Stn.
Sample CHGSS 129.
Pleistocene F-W. dune system persist little
altered. North of the Murray River and Westof the Darling River, the terrain consists of the
Pliocene/ Pleistocene Blanchetown Clay with
discontinuous surlieial sand. A 'steady state
relief has been established.
Palaeontology
As most of the outcropping formations were
deposited under semi-arid conditions, plant
fossils are not common. The highly oxidizing
environment (see Maher, this Memoir for cli-
matology) is inimical to the preservation of
plants, The ?water plants in the Blanchetown
Clay at 'Warrakoo' have already been men-tioned, and the fossil trees at the Chowilla
Dam site. Blackburn (1971) has recorded a
podocarp root from Telangatuk East, further
S. in Victoria. Gill and Bethune (1967) ob-
tained pieces of wood from a bore at Narran-
dera that dated 32,000 ± 1,200 yr and >33,000 yr. Wood was obtained by Mr Betlumein a bore at Raak, Victoria. At Mildura woodfrom a depth of 10 7 m near the Psyche Bendpumping station dated 5,400 yr (NZ-196).
Churchill (this Memoir) has recorded pollens
and spores. The amount was small but further
work in this field is projected because this flora
marks the incoming of the semi-aridity. Near
where the sample was collected for DrChurchill, a thicker section was later revealed
by removal o\' slope wash. This outcrop pre-
sented I m of black sediments, providing scope
for serial study. The cliffs of Salt Creek at the
S. end consist of Chowilla Sand with slumpmaterial forming a talus slope at the base that
probably conceals Parilla Sand. The palynolo-
gical site is where the Blanchetown Clay begins
to lens in within the Chowilla Sand. The deposit
is that of a lake shore.
The Animalia arc better represented amongthe fossils. Marshall's paper in this Memoirdescribes the vertebrates, which can be com-pared with those from Lake Tandou in the sameregion (Merrilees, this Memoir).
The formations described have the following
fossils:
1. Bookpurnong Formation, Cheltenhamianmarine fossils of shallow water facies.
2. Parilla Sand. Samples of these floodplain
sediments yielded no spores or other fossils,
but as the beds arc unoxidized and carbona-
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 77
ceous in some places, further search may find
them.
3. Moorna Formation. Fish (including Neo-
ceratodus) , tortoise, ratite bird (egg shell)
,
marsupials and ostracods.
4. Blanchetown Clay (and its limestone mem-bers). Fish, marsupials, worms (casts), mol-
luscs and ostracods (McKenzie and Gill 1968).
5. Chowilla Sand. Fish, tortoise, marsupials,
crustaceans (gastroliths) and ostracods. AtBone Gulch on Moorna Station, in the Chowilla
Sand below the Blanchetown Clay are struc-
tures interpreted as burrows (PL 7).
6. Rufiis Formation. Fish, ratite bird (egg
shell), marsupials and rats. In Browns Paddock,
Keera Station (Fig. 38), are 'fossil' formicaria.
Noting a circular patch of carbonate rising
10 cm above the surface of the carbonate-free
red sand hardpan, I sank an auger to investigate
the anomaly, and found that a carbonate rich
zone continued to a depth of over 50 cm. Thestructure was then excavated by spade, and
recognized by K. N. G. Simpson as the former
nest of a large ant such as the Bulldog Ant(Myrmecobius) . The galleries attained a maxi-
mal diameter of 13 cm. The carbonate at the
surface had been carried up by the ants from
the B horizon of the soil, including pieces upto 12 cm long. We found a small species of
ant no larger than 10 mm in length occupying
the galleries in the top 2-5 cm of the aban-
doned formicarium.
7. Nulla Nulla Sand. Molluscs, reptiles, mar-
supials (including extinct genera and species),
rats and man (skeletons and middens).
8. Talgarry Sand. Same range of fossils, but
all living species.
9. Monoman Formation. Fossil wood and
marsupials.
10. Coonambidgal Formation. Aboriginal
burials and middens.
Aboriginal bones and middens along with some
marsupial bones were found in the hamada on
the W. side of Lake Victoria. As our investiga-
tion was only at reconnaissance level, there are
probably many more palaeontological sites to
be found. Certainly more fossils will come
from a systematic treatment of the sites dis-
covered.
i \ 26 125 'M3 i$ -.! u ;
. van
Fig. 61—Histograms 92-93 of grain
size analysis of Talgarry Sandmember of the Lake Victoria
Sand formation.92 ± Sample CHGGS 1 12 at
BM 1, Talgarry Stn. just S.
of boundary fence shown in
(Fig. 16. 93 = Matrix of
Skeleton A from Gulch SI
(Fig. 16), Talgarry Stn., E.
side of Lake Victoria, N.S.W.Sample CHG 309.
Following very heavy rain, bones in the
Nulla Nulla Sand can achieve consistency
rather like that of cheese, and can be readily
cut with a knife. On drying out, they become
hard again. While plastic they can be deformed.
A few workers in other countries have noted
(pers. coram.) the same or a similar phenom-
enon, but no study of it has been discovered
in the literature.
Palaeoclimatology
Je nien vais chercher un grand peutetre.
Rabelais
This subject is as difficult as it is important.
For the former reason many avoid it. Adequate
treatment involves great masses of detail plus
unifying ideas. However, if all authors pro-
vided such palaeoclimatic data as they have,
and passed on any ideas that occurred to them,
we would be much further advanced in this
large field of learning. The palaeoclimatology
of the study area is approached in this way.
Terrestrial palaeoclimatology is more difficult
than marine palaeoclimatology, because morevariables exist. Local microclimates are pre-
78 EDMUND D. GILL
Figure 62
Sedimentology
The Sorting Coefficient (So) is of sediment with-
out carbonate or clay, unless otherwise stated.
Perfect sorting » l t2.5 = well sorted, 3
normally sorted, and 4.5 - poorly sorted
(Trask.
)
o--*
3
0)•—*
as03
CO
aC
XIUaV*
PARI LI
t—
I
U
JL SANJ
oCO
3
i 78 2.92 1.30 Floodplain
2 115 13.86 1. 17
MOORNA FORMATION
3 113
111
2 . 18 27.65 1.54
4 17 . 9 3.42 1.49 Dolomite removed
in2from sand grains
5 - - 1.36 Untreated sample;
channel facies
6 118 12 .2G 0. 56 1.25
7 94 19.62
8 108 2.88 1.45
17
19
20
21
22
23
24
25
26
CHOWILLA SAND
9 38a 0. 85 1 23 Channel facit
10 77 1. 98 1 . 14
1 1 109 3.66 11. 87 1 4812 117 3. 20 0. 51 1 57
13 41I
Early tests;
14 431 washings not
weighed15 116 4. 39 1 52
16 114 18. 16 1 .68
BLANCHKTOWN CLAY
42 68
18 151a 21. 95 18.35 1 .27
Normal facies;
more clay some-timesSandy facies with
Corbicula
WOORINEN FORMATION
32
33
34
35*10.
35?
36 3.
37 11.
3£b
o
02
38
98
29
87
40
72
01
88
9
70
92
34
30
34
29
29
30
52
1.31
Dune facies
Repeat
Floodplain or lake
of underlying
terrain
27 1 18 30. 14 2.34 1 . 37
<i RUFUS ALLUVIUMa
u
OCOp-H
X *
28'
rHaa03
CO
74
oj
ao£2u03
u
13. 73
r-H
u
8. 01
oLO
1.59 Floodplain facies
Outlier L Auger1
29 75 14.33 8. 23 1.58
30 76 10. 71 26. 15 1.52
31 77 7. 51 14.51 1.48
32 78 5. 27 10. 77 1.37
33 79 3. 33 6. 80 1. 36
34 80 5. 16 4.40 1.50
35 81 2.66 15.44 1.37
36 82 2.07 42.08 1.37
37 83 1. 57 68. 23 1.65
38 84 22. 97 1.46
39 85 18.61 1.69 Outlier 1, Auger2
40 86 7. 14 20. 95 1.60
41 87 5.30 15.67 1.54
42 88 4. 37 15. 86 1. 55
43 89 2.03 8.34 1.57
44 91 8.47 1.57
45 100 22.47 1.40 Outlier 1, Auger3
46 101 4.62 1. 20
47 92 2. 93 1.57 Outlier 1, Auger4
48 93 5.93 54. 28 1.54
49 94 8. 98 25. 16 1.31
50 95 5. 59 1.29
51 96 4.65 19.63 1.56
52 97 8. 88 1.33
53 98 29.57 14. 75
54 102 6.82 59. 98 1.72 Outlier 2, Auger1
55 104 3.40 18. 37 1.28
56 105 0.61 2. 38 1.29
57 106 18.97 0. 50 1 .37
58 107 30. 09 9.62 1. 51
59 400 4.40 2. 26 Site 1, Auger 1
60 401 1.6 12. 9
61 402 12.7 16.0 1.78
6 2 403 63. 9 2.5563 4 04 47. 5 4.364 405 18.6 1.61 Site 1, Auger 4
65 406 16. 8 35. 7 2.0066 407 5. 3 39.08 2.0467 408 42.07 1. 57
68 409 0. 93 3.80 1.2869 410 12.0 1 .41 Site 1, Auger 1
70 411 6.4 1.4171 412 4. 98 1.43
72 413 4.0 1.45
73 414 16.2 1.46
74 415 69.2 1.41 Site 6, Auger 2
75 416 9.69 12. 5 2.22
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 79
s
uO
03
a.
s
CO
g
In
U
76 417 2. 11 2.62 1.52
77 418 12.6 1.43
78 419 18.6 1.50
79 420 4.76 1.42
80 421 3.05 1.33
81 422 1.60 1.33
82 423 2.25 1.06
MONOMAN FORMATION
83 55
COONAMBIDGAL FORMATION
84 215 18.80 9.98 2. 10 Lake Wallawalladoubtfully
referred this
formation
85 31 25.80 1.22
86 95 1.68 15.5 1.34
87 39 11.61 1.43
88 40 3.70 1.46
NULLA SAND
89 110 1.09 4.18 1.53
90 98 7.42 6.53 1.60
91 129 2.37 0.12 1.56
TALGARRY SAND
92 309 2.30 1.44
93 112 3.48 1.77
valent. Living organisms have inbuilt adaptive
mechanisms so that within limits they change
with the changing environment. Climacteric
events are sometimes of sufficient magnitude
to leave a record of these atypical happenings,
diverting attention from the more significant
persisting characters of the terrain. The solu-
tion to this problem in our investigation was
to consider evidence that was consistent over
large areas of country. For example, the E-W.
longitudinal dune system, once mobile but now
immobile as a system, covers some 200,000
km2. This extensive regional change in the
status of sediments sensitive to effective rain-
fall and wind regime is a matter of prime
palaeoclimatic importance. Another example
is that beneath this semi-arid terrain, clogged
with surficial carbonates and sulphates (proof
of low humidity and little leaching), there lies
a fossil lateritic profile (proof of high humidity
and very deep leaching). Smaller changes of
climate are harder to discern, but they are best
deciphered in marginal areas where maximalchange has occurred. Many misinterpretations
of past climates have been due to lack of an
adequate chronology. In the present study, the
order of events is unequivocal, and progress
has been made with chronometric measure-
ments, but a great deal remains to be done.
Thus, while the change from humidity to semi-
aridity can be located in the stratigraphic suc-
cession, the time cannot be defined with any
precision.
In SE. Australia, the palaeoclimatic history
prior to the period represented by the rocks
exposed in the study area may be generalized
as follows:
1. Cretaceous Frigid Supercontinental Climate
It is widely held that Australia was at-
tached to the Antarctic continent in the Creta-
ceous (Smith and Hallam 1970), that the south
pole was nearby (Wellman et al. 1969), but
higher temperature zones existed to the NW.(Gill 1972d and references), and that Australia
drifted away from Antarctica in the UpperCretaceous to Eocene (Veevers et ah 1971,
Jones 1971). The high percentage of felspar
(,— 40%) in the extensive Lower Cretaceous
rocks of Victoria has long been a puzzle, but
if the climate were frigid, this can be under-
stood. Solitary pebbles of plutonic and meta-
morphic rocks and patches of gravel unassocia-
ted with current bedding, and not streamed out
in lenses, may be drop pebbles from floating
ice. In 1949 I observed with Mr A. A. Baker
and others a cliff in the Cape Patterson area
in S. Gippsland, E. Victoria, a large mass of
laminated fine arkose about 4*5 m long with
sharp broken edges, and fragments 'floating'
in the surrounding matrix. This massive piece
of rock was certainly not transported by water.
It is completely out of character with the
surrounding sediments and their dynamics; its
edges are uneroded. At the time, the only solu-
tion I could conceive was that it fell from a
80 EDMUND D. GILL
cliff, but this could not be accepted by reason
of the stratigraphy shown in the wide shore
platforms and high cliffs. To explain this as
a rock mass dropped from melting ice into the
sediments on the floor of a lake explains all
the factors observed. More recently, varves have
been found at Koonwarra, with fossils that
appear to represent winterkill (Waldman 1971).
2. Mid-Tertiary Subtropical or Tropical
Rainforest Climate
The marine biota is a warm one, while onland there were large reptiles and thick rain-
forests that formed extensive brown coals (Gill
1961a-b, Duigan 1966, Dawson and Sneddon
1969, Dawson 1970). At first, the climate
appears to have been without seasonal change
in humidity, so that deep leaching with kaolini-
zation occurred to 60 m. These conditions
created the Nunawading Terrain (Gill 1964)and were succeeded by a seasonal monsoonal
type climate that yielded laterite. It is significant
that at that time (Pliocene, Gill 1971) laterite
developed both land- and seaward of the
Dividing Range. The type section of Chelten-
hamian marine sediments on the coast at
Melbourne is lateritized, as also are sediments
overlying the Cheltcnhamian fossil bed at
Tareena Station on the Murray River. In someplaces, the Nunawading Terrain is overlain bylater sediments that are lateritized.
If at this time, SE. Australia was drifting N.from the frigid zone towards its present tem-
perate zone position, the tropical or subtropical
climate must have been due to a change in
world climate that widened the tropical belt.
On what is known of world climates, this could
well be.
Axelrod and Bailey (1969) have suggested
that climatic change in the Miocene could havebeen effected simply by transgressive seas re-
ducing extremes, 'and hence biota (both marineand non-marine) would assume a somewhatwarmer aspect under the more equable condi-
tions'. However, more than a change in equa-bility is involved in SE. Australia as is provedby gross change in soil types, by the presence
of certain stenothermal genera, and such majorshifts. The change from the cold climate in early
Tertiary to more or less tropical conditions in
the mid-Tertiary, followed by a major change
in fauna and flora with cooling and drying cli-
mate in the upper Tertiary is more than can be
explained by an increase in equability. Thechange took place both N. and S. of the Di-
viding Range. The sea exercises little control of
the climate N. of the Range. It should be noted
that Australia is naturally more equable in
climate than most continents (e.g. Europe, Asia
and N. America) because it is set in a wideexpanse of ocean and possesses no land link
with the Antarctic. Jn addition, its mountains
are exceptionally low, for Australia is the flat-
test continent.
The deep kaolinization of the NunawadingTerrain is apparent in the solid rock areas of
Victoria, New South Wales and South Australia,
but not of course in the rapidly sinking MurrayBasin. Moreover, the study area was invaded
by the sea during the mid-Tertiary.
Following the above two palaeoclimates,
there were three phases that left their impress
on the terrain by significantly different types
of palcosols, viz.
(a) A lateritic profile that accumulated non-magnetic iron oxide.
(b) A paleosol that accumulated silica in sandsand common opal in clays. Such a soil
infers a different pH from lateritizing con-
ditions.
(c) A series of pedocals that accumulatedearthy carbonates to crystalline nodules,
the latter typically with an amorphouscentre but concentric laminae on the out-
side.
3. Monsoonal type climate
This develops rainforest, but is characterized
by alternating wet and dry seasons. Heavy rain-
fall drenches the earth with copious warmwaters that leach it deeply, causing kaoliniza-
tion. This is followed by a dry season whenthe water table falls and the air enters, oxidizing
the zone above the table. Mottles of iron oxideform, and if the process continues long enough,the part of the mottled zone always exposed to
oxidation becomes fully oxidized, resulting in
the ironstone horizon which is laterite' in the
strictest sense.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 81
The lateritic profile is so deep and the leach-
ing so intense that it reflects a climate which is
the converse of the present semi-arid conditions,
wherein soluble carbonates remain at the sur-
face. The next climate represents the transition
between these extremes.
4. Monsoon/Arid Transition Climate
Whether this is an adequate characterization
of this climate, there is not enough evidence to
say, but it is established that a climate causing
deep leaching changed to a desertic one. Thepresent climate is semi-arid but when the EW.dunes were active, the climate would be classi-
fiable as arid. This transition is reflected in the
sediments, the fossils and the soils. Tectonic
movements about this time appear to have
further uplifted the Dividing Range, which
event must also have contributed to change of
climate in the areas concerned. However, the
climatic change extends far beyond the influ-
ence of the Dividing Range (e.g. Lowry 1970,
Twidale 1972), so factors other than uplift
are involved.
When Australian Oil and Gas sank Lake
Victoria No. 1 well, they met late Miocene
marine Bookpurnong Beds at 308 feet (94 m).
Above that was a sequence of gypseous sands
(arid/semi-arid) passing into pyritic sands
(pluvial) with carbonized wood fragments and
fish bones. The passage is from alkaline to
acidic sediments.
The lowest formation outcropping along the
Murray River in the study area is Parilla Sand
which is a gray clayey fine sand. It is riverine,
and its carbon content (high in places) con-
trasts with the low carbon content of the
Coonambidgal Formation which was deposited
by the same river. It is succeeded by the well
washed oxidized Chowilla Sand (channel de-
posit) which interdigitates freely with the
Blanchetown Clay. For the present purpose the
latter two formations can be treated as one,
and then the successively higher levels will
represent successively younger zones. The
Blanchetown Clay is thickest in the NampooStation cliffs, and it is at the base of the
exposure that evidence of the incoming drier
climate is first met. By tracing the river cliffs
to the NW., the gradual replacement of
Blanchetown Clay by Chowilla Sand can befollowed through Cal Lai to the Kulcurna Sta-
tion homestead, where Blanchetown Clay is
raising. Nearby at the S. end of Salt Creek the
formation lenses in again well up in the
Chowilla Sand section, i.e. at what is interpreted
to be a later horizon. There the present semi-
arid flora has been recognized by Churchill
(this Memoir).At Sharp Point on the Nampoo cliffs (Fig.
26), three types of evidence combine to in-
dicate the increasing dryness of the climate:
(a) Deposition of dolomite (cf. Zenger 1972).As indicated elsewhere in this paper, it is con-
sidered that this rock was a product of warmtemperatures, low rainfall and high evapora-
tion resulting in saline waters. Such dolomite is
precipitated in the Coorong in S. Australia at
the present time. The two dolomite layers showrepetition of the sensitive conditions necessary
for deposition. Deposits also occur at Triple
Swamp (Fig. 24) and elsewhere low in the
Blanchetown Clay/Chowilla Sand complex. Thetop of this complex is at plateau level. Nodolomite has been found in the top three
quarters of the complex.
(b) Drying up of Lake. The BlanchetownClay is a lacustrine deposit. To spread channel
sands (Chowilla Sand) across a lake bed(Blanchetown Clay) is evidence of shallowing,
as also is the deposition of the dolomite. Thelake dried up eventually as is proved by the
formation of a soil and development of rhizo-
morphs. This soil accumulated silica, which at
Sharp Point is in the form of common opal,
but at Cal Lai is a cement forming silicified
sandstone. In places, e.g. N. Salt Creek, silcrete
occurs. I followed opal deposited in clay to
where in the sandy facies nearby it becamesilicified sand to silcrete.
Deposition of opal appears to be controlled
by impedance of drainage, and so presumablythe formation of a gel. The formation of opal
has been attributed to different geological
periods, including the time of development of
the lateritic profile. Because of basin sinking,
the succession is here extended, making it
easier to distinguish successive geologic events.
In this instance, it can be proved that the opal
is younger than the laterite. The upper opal
82 EDMUND D. GILL
layer at Sharp Point is a cemented breccia in
places, suggesting some kind of dessication
followed by further deposition.
This paleosol is very widespread. It is the
Karoonda Surface or Terrain of Firman (1967).
For example, Kosciusko Epoch sediments over-
lying kaolinized granitic rocks at Home Rule,
N.S.W., are silicified, and this may be referable
to the same climatic phase. Silicification has
been noted near the surface in many places in
the Mallee area of N. Victoria. It seems a good
thing to give the same name to the fossil terrain
and the paleosol that occurred on it. In this
case, the latter could be called the KaroondaPedoderm. No application has been made for
the recognition of this as a stratigraphic name.
The paleosol is often eroded, as can be seen
(for example) by following the silcrete layer
along the Merbein Cliffs (Fig. 36). It is absent
in other places, although features such as
mottling of the Blanchetown Clay (e.g. in the
S. bank of the Murray Cliff at Yelta) may bean expression of it.
In W. Victoria a fossiliferous site has beendescribed which records the changeover period
from the humid climate with a coniier/Notho-fagus flora to the present dry Eucalyptus/
Acacia flora (Gill 1957). A vertebrate faunahas been described from the same horizon
(Turnbull and Lundelius 1970). A basalt flow4- 35 m.y. old has sealed off the deposit. Fluvia-
tile sediments under thick basalts at Smeatonare referable to this same changeover period(Gill, in prep.). Dr Ian McDougall kindly
undertook to date for me a sample from oneof the overlying basalt flows with the result of
211 m.y. (Aziz-ur-Rahman and McDougall1972). The changeover time appears to havebeen of considerable duration, and so is listed
here as a separate palaeoclimatic phase.
5. Arid/Semi-arid Climate
All formations later than the Chowilla Sand/Blanchetown Clay complex have carbonate
associated with them, except for those formingthe present river floodplain. Polygenetic car-
bonate soils characterize the surface of the
Blanchetown Clay forming the flat plateau
country. At least three cycles are involved, as
represented by:
(a) Crystalline calcite nodules included in a
younger generation of such nodules.
(b) The younger generation of these hard
nodules which are r~> 28,000 years old.
(c) The still younger accumulation of earthy
carbonate r^ 16,000 years old.
In the dunefields, corresponding periods of
terrain stability and instability can be traced
over extensive areas. In the erosion amphi-
theatres on the W. side of Lake Victoria there
are matching periods of erosion and deposition,
judging by the few radiocarbon dates available.
There are indications also of Holocene oscil-
lations, finishing with the 'indurated layer' (Fig.
46) currently being deposited.
Mid-Holocene Thermal Maximum. Such is de-
scribed from many parts of the world, but manyAustralian workers have claimed it does not
occur here. This is too much to claim as so
little work has been done on palaeoclimatology.
However, one can describe the evidence onefinds, and offer an interpretation of it.
Before CI 4 and U/Th dating, evidence of
more than one drier period was telescoped into
a single 'Arid Period' believed to be of mid-Holocene age. In 1955 the author wrote a paperplacing the term 'Arid Period' (as applied to
the mid-Holocene ) in inverted commasand stating it was a misnomer. It was pointed
out that the basic climatic change is one of
mean temperature, that this does not neces-
sarily mean change in humidity in a particular
direction, and therefore the term 'Arid Period'
should not be used. In spite of this, numerousauthors have referred to this paper as indicating
my belief in the mid-Holocene 'Arid Period',
and it is hoped that this note will encouragethe accurate reading of what was written.
Nevertheless, I do think that in Australiathere was a small but definite mid-Holoceneclimatic oscillation characterized by a smallrise in temperature such as has been describedfrom many parts of the world. In the studyarea the following evidence has been noted:1. A disconformity between the Monoman and
Coonambidgal formations at the ChowillaDam site is marked by a layer of fossil
stumps and logs that range in age from4,080 to 7,200 years. The sediments beloware oxidized.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 83
2. A disconformity exists in the large lunette
on the E. side of Lake Victoria betweenthe Nulla Nulla Sand Member and theTalgarry Sand Member. Aboriginal middenshells in the paleosol marking the discon-formity dated 6,360 years. At the end ofthis period, gulches were excavated by winderosion then later infilled with TalgarrySand.
3. On the E. side of Dolomite Gulch at Triple
Swamp on Moorna Station an Aboriginal
midden was found between the A and Bhorizons of the soil. The original A horizon
must therefore have been eroded away andlater reconstituted. The date of the midden(7,210 yr) is within the erosion interval.
At Port Campbell in Western Victoria awinnowing of the A horizon of a humuspodsol was proved, and the interval as
represented by charcoal samples was 4,830to 5,700 years. Limits on the period of
time involved are placed by other sampleswhich dated 7,380 and 3,880 years respec-
tively (Gill 1965). Along the coast of Vic-
toria are emerged marine shell beds with
evidence of slight warming that give dates
from 3,980 to 7,040 years (Gill and Hopley1972).
4. In N. Victoria a ridge of granite (Terricks
Range) stands up through the sediments
of the Murray Basin. At Mitiamo, the ex-
cavation of the clayey sand on a pedimentrevealed an Aboriginal skeleton at the base
that dated 5,540 yr. B.P. so this was pre-
sumably in a time of deflation when the
pediment was swept almost bare (Gill
1967).
5. The most characteristic mammal of this
semi-arid country is the Red KangarooMegaleia rufa. It does not occur now S.
of the Dividing Range in Victoria. How-ever, in the mid-Holocene it was in S.
Victoria. Mr E. H. Wilkinson and Dr P.
Lang found it at the outlet of Lake Gnar-
purt in Western Victoria. The skeleton wasin fine lacustrine sediments with Coxiella
shells, which dated 4,550 years (Gill
1971b). These thin-shelled brackish-water
gasteropods give a date on the young side
where check has been possible, so the Red
Kangaroo may be a little older, but still
mid-Holocene.
Early records and surface bones showthat Bettongia leseur was also characteristic
of the semi-arid fauna before the introduc-
tion of European cats and dogs. Bones of
this species occur at Bushfield in W. Vic-
toria, where they underlie the Tower Hill
Tuff dated 7,300 years B.P. (GUI 1963,1972c, Wakefield 1972).
6. The Last 30,000 Years
Radiocarbon dating provides reasonable
chronologic control for this period. Insufficient
is known to make a proper palaeoclimatic chart,
and too few dates have been obtained. Thecarbonate dates in the upper part of the radio-
carbon range need to be treated with reserve
until they are better understood. However, the
following table is the picture that appears onpresent evidence, and it is presented as a stimu-
lus to further thought. Some of the contem-porary events S. of the Dividing Range are also
given.
Order of
C14 Age Events
0-4,000 Emplacement of youngestyears B.P. colluvial fans, dunes, and
floodplain deposits (Coonam-bidgal Formation, DunedinPark Sand Member, etc).
Formation of youngest car-
bonate soils, and youngest
Aboriginal middens andovens.
Modern erosion (0-150 yr.)
Probably two phases in this
period.
4,000-6,500 Terrain instability in study
area.
In Lake Victoria lunette, soil
(with midden dated 6,360 y.)cut by gulches and covered byTalgarry Sand Member.Extant vertebrate fauna with
minor changes. No surficial
Aboriginal sites of this age
found, but burials of this
period prove their presence.
84 EDMUND D. GILL
6,500-9,000
9,000-15,500
15,500-17,500
Disconformity between Coo-
nambidgal Formation and
Monoman Formation at
Chowilla Dam site marked
by fossil trees, and oxidation
of underlying sediments.
Pediment with Aboriginal
skeleton buried at Mitiamo,
N. Victoria.
Deposition in Devon Downssite further down Murray R.
Deposition of pedogenic car-
bonate N. and S. of Dividing
Range (not presently so de-
posited S. of Range).
Red Kangaroo migrated S. of
Dividing Range.
Deflation of topsoil at Port
Campbell, S. of Range, wheresurface presently well stabil-
ized.
Higher sea level on coast of
Victoria shown by emergedmarine platforms, shellbeds,
and other shoreline structures.
Terrain stability.
This period is ill-defined, but
before the above sand move-ments began, a red soil wasformed over most if not all of
the large Lake Victoria lun-
ette. Below this soil, high in
the Nulla Nulla Sand were
bones of an extinct kangaroo
that dated 9,610 yr.
Terrain instability.
Deposition of upper part of
Nulla Nulla Sand Member in
the Lake Victoria lunette.
Deposition of the uppermostmember of the E-W. dunesystem.
Extinct vertebrates.
Terrain stability.
Paleosol formed in Lake Vic-
toria lunette, and in E-W.dune system (Woorinen For-
mation).
Paleosol in clay (parna) dune
17,500-27,500
27,500-30,000
on E. side of Lake Coranga-
mite, S. of Dividing Range.
Terrain instability.
Penultimate member of E-W.dune system cmplaced.
Deposition of lower part of
Nulla Nulla Sand in Lake
Victoria lunette.
Extinct marsupials present.
In some places a paleosol in-
tervenes at 24,000 yr. (e.g.
Nyah, W.).
Terrain stability.
Paleosols formed in E-W.dune system.
Probably Lake Victoria lun-
ette initiated in this period.
Pedogenic calcite nodules with
laminated cortices below E-W. dune-field and on general
terrain of Blanchetown Clay.
At least two generations are
involved, the second extend-
ing back beyond the range of
radiocarbon dating.
S. of Dividing Range, shells
from high level beach of LakeCorangamite dated 28,240 yr
(Gill 1971b).
The information available so far indicates
alternating periods of terrain stability and in-
stability. These are judged to be a function of
significant variations in dryness, and probablyalso in windiness (Wilson and Hendy 1971).In the more humid country S. of the Dividing
Range there is a similar pattern but, as maybe expected, the time limits of the periods are
not exactly the same.
Ecology
Two ecologic zones characterize this country
(1) the flatland of the general terrain, and (2)the flatland of the wide floodplain incised 37 mtherein.
1. In the study area, the general terrain is
markedly different N. and S. of the MurrayRiver.
(a) To the S. the country is essentially
sandy due to the dunes and sand
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 85
spreads of the Woorinen Formation.
This is the Millewa Land System of
Rowan and Downes (1963). Mallee
scrub is the dominant vegetation.
(b) To the N. of the river the country is
essentially clayey (due to the Blanche-
town Clay substrate) with only a
veneer of sand. Saltbush is the domin-
ant vegetation.
The sandy facies to the S. of the Murray R.
is an expression of fluviatile activity later modi-
fied by wind action. The clayey substrate to the
N. of the river is an expression of lacustrine
activity. The River Murray has followed the
interface between these blocks of different sedi-
ments. The sands are much easier to erode than
the clays. These generalizations provide a basic
understanding of the country, but the actual
situation is of course much more complex. Thehypothesis put forward here for the course of
the Murray requires testing by more detailed
investigation. A tectonic factor is also likely.
2. The floodplain has been divided by Rowanand Downes (1963) into two zones:
(a) Ned's Corner Land System consisting
of higher ground (including the Rufus
Formation). The dominant vegetation
is saltbush.
(b) Lindsay Island Land System, which is
Holocene floodplain (Coonambidgal
Formation) with channel border dunes
and such. The dominant vegetation is
Eucalyptus camaldulensis lining the
streams, with E. larglflorens on the
floodplains behind.
Information on the present fauna is given by
Simpson (this Memoir) and that of the past
by Marshall and Merrilees (this Memoir) plus
the paleontology section of this paper. Aninvestigation of the operation of the ecosystems
involved is needed.
During the drought of 1967-8 many sheep
died of malnutrition. Although in good seasons
there is only minor competition between sheep
and kangaroos, it becomes direct in time of
drought. Graziers culled the kangaroo popula-
tion, but the latter were still numerous, as they
came from inland to within watering distance of
the river. As soon as the rains came, the kanga-
roos disappeared inland. During the drought
some birds that are normally further N. also
came S. Thus at such a time there is a tem-
porary population zone shift.
The Aboriginals were oriented to the river
in a similar way, and went into the hinterland
when the rains formed pools and re-charged
small reservoirs in bodies of sand underlain byclay. Their mussel middens, implements, and
fireplaces provide the evidence. As mussels
occur only in the rivers and lakes, they musthave been carried inland. They were noted onTalgarry Station nine miles E. of Lake Victoria,
the nearest source.
Evolution of Ecosystem
When the Lower Cretaceous epicontinental
seas withdrew, Australia changed from a pattern
of islands to its present island continent status.
The river systems ancestral to the present ones
were then established. At that time SE. Aus-tralia had a subantarctic climate which graded
to a subtropical one in the N. (Gill 1972d).
This gradually changed until in mid-Tertiary
a subtropical climate with extensive rainforests
(that formed brown coals in grabens, and deep-
leached soil profiles on horsts) existed in this
area. From then till now there has been an
overall drop in temperature and humidity. Theincoming of the present aridity to the study
area is described elsewhere in this paper. Thetime was ***> 2 m.y. ago. In that period the
fauna in and around the river adjusted to a
regime where the river could be anything from
a string of salty pools in drought to a flood
reaching tens of kilometres wide. Thus great
extremes of chemical and physical conditions
were involved. When the warm flood waters
crept across the extensive floodplains, a wholenew pulse of life occurred that resulted in plant
renewal and the triggering of a complex faunal
cycle. The millions of microzoa fed the small
fish that emerged, and on which birds andother animals fed (Lake 1967). The intro-
duction of dams and weirs has interfered
with this natural cycle, but some simulation
of it by dam control would be possible.
Only 1 9 species of native freshwater fish
live in the vast Murray/Darling R. system, but
they show some remarkable adaptions to their
86 EDMUND D. GILL
unusual habitat. These must have evolved since
the present semi-arid climate came. Eight spe-
cies of fish have been introduced. Rainbow andbrown trout do well in the colder upper reaches
of the rivers, while redfin and tench do well in
the dams.
Conservation of the Ecosystem
The present ecosystem took 2 m.y. to
develop, and is very finely adjusted to very un-
usual conditions. In that period of time manynew species have appeared and many becameextinct in the study area. The lungfish (Neo-ceradotus) and many marsupials have disap-
peared from it (Marshall and Merrilees, this
Memoir). On the other hand, a fossil tortoise
found in Fisherman's Cliff, Moorna Station, in
Chowilla Sand between the Moorna Formationand the Blanchetown Clay could not be dis-
tinguished by Professor J. W. Warren from its
living counterpart.
The dry terrain is subject to erosion, but wasmuch more stable before the introduction of
herds of sheep and other animals. To conservethe ecosystem requires careful management.The first two years of our study (1967-8) weredrought years, and wind erosion was very ex-
tensive. However, the incidence of erosion
varied from station to station. The examinationof the air photos shows the unused or little usedareas are much more stable. On the other hand,some areas that were badly eroded have beenrestored by good management.
The extensive occurrence of Aboriginal mid-dens with charcoal in the topsoil extendingback to 2,000 yr ago shows that until recently
the surface has kept intact over wide areas for
two millenia. As charcoal is one of the mosteasily eroded materials, its presence is goodevidence of undisturbed ground. On the W.side of Lake Victoria, the stratigraphy of thegulches and colluvial fans shows that in thepast there have been periods of erosion andperiods of infilling (Fig. 47). To delineate
these and date them would be a useful project.
Active gullying is now in progress, and there
is difference of opinion as to when this wasinitiated. The latest cementation by carbonateis the fixing of the top 10-15 cm of the latest
colluvial fans. This indurated layer contains
middens, Aboriginal bones, and other dateable
materials. A midden of this layer on Noola
Station contained carbonized twigs (probably
saltbush) that gave a modern reading uponradiocarbon assay. This means that the age
is 200 yr at most, so the erosion has occurred
since the arrival of Europeans, their sheep andrabbits and their machines.
Water erosion following heavy rain is assis-
ted by the nature of the spoor of sheep andother ungulates. Kangaroos and emus havediscontinuous spoor, but sheep scuff the groundwith their hooves and excavate a rut downwhich water rushes, so initiating a gulch. Simi-
larly I have seen rain following the tyre marksof a motor car. Thus the terrain is sensitive to
wind and water erosion, necessitating careful
management, if it is to be conserved. This is
desirable both economically and scientifically.
The country has a considerable capacity for
recovery.
Aboriginal Ecology
'The earth's surface being to them a bookthey always read.' Major Mitchell
Having been over 30,000 yr in the country(Bowler**/ al, 1970) the Australian Aboriginalshad long become an integral part of the eco-system by the time Europeans arrived. TheAboriginals were one with their country. It
was in character for them to prefer the local
opal for their implements rather than the moreefficient imported flint 'because it was part ofthem' (so said Aboriginals in the WesternDesert). Natives on the Darling River askedMitchell to return water he had taken from it.
They had totemic relationships with animalsin the local fauna. During the long period theyinhabited this area, there were changes of cli-
mate and many of the local species becameextinct. They were a conservative people (keep-ing to the ancient core/flake culture), butmanaged to accommodate themselves to thesechanges. When animals that became extinctwere totems, deep-seated mental adjustmentswere necessary. Economic change was enforcedby loss of technological raw materials (bone,skin, hair, tendons, fat, etc.) from the animalsthat went extinct, plus loss of a food source.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 87
m m H=a»- -fx aFig. 63—Histogram of weights of stones from Aboriginal oven in Scrub Paddock at top of hill SW.
of Dickie's Gate, and E. of Salt Creek, Kulcurna Station, N.S.W. Site 2. Categories are of 10grams each, extending from 1-10 to 501-510; then there is a gap, and the final category is
581-590.
Two sources provide information on the wayof life of the Aboriginals of this area:
1
.
Early explorers and settlers, Mitchell
(1839), for example, described the Darling R.
natives, their dress, artefacts (spears, canoes)
and something of their daily activities such
as hunting marsupials, spearing fish, netting
ducks, collecting mussels and roots, using fire,
dancing, communicating by calls and fire-
signals, and burying the dead. Sturt (1849)described the natives at Lake Victoria andthereabouts. The notes of Beveridge (1865),Hawdon (1852), Morton (1861) and others
are likewise helpful. Lawrence (1968) has
accumulated valuable background information.
2. Archaeologic evidence. In this Memoir Casey
has described artefacts, Blackwood and Simpsonthe Aboriginal skeletons, and Sandison their
pathology. Many middens have been recorded.
From an ecologic point of view, three types of
Aboriginal fireplace need to be distinguished:
(a) The midden. In this area these consist of
mussel shells (Velesunio), charcoal, ash and
sometimes burnt ground (Gill 1969). The term
is derived from a Scandinavian word for a
refuse heap, and so is appropriate.
(b) The oven. This is a fireplace characterized
by cooking stones. These are accompanied bycharcoal, ash and sometimes burnt ground.
Due to the shortage of rocks in this area, all
manner of materials were used. Sandstone,
ironstone and pieces of wasps1
nests were uti-
lized. Over much of the area, only pedogenic
materials were available, such as carbonate
nodules and silcrete. Figure 63 shows the size
distribution of carbonate nodules in an oven
on Kulcurna Station illustrated in Gill 1969,
p. 181. The weight analysis is of the oven stones
shown in the well-defined round oven area in
the foreground of the photograph.
Where no stones were available for ovens,
the Aboriginals baked mud lumps for this pur-
pose. It is the nearest they came to makingpottery. Impressions that appear to have beenmade with fingers, or in one instance an arm,have been noted. Barbetti (this volume) hasstudied the palaeomagnetism of these ovenclays. Thermoluminescent tests have also beenmade on them. See PI. 10, fig. 3.
In some ovens the stones were a mixture of
imported rocks and the baked clay lumps.
Mitchell (1839, vol. 2, p. 81) states' 'Thecommon process of natives ... is to lay the
food between layers of heated stones'. Bev-eridge (1865) described the customs of the
Murray R. natives among whom he lived in
the Lake Boga area. He said. 'They cook their
food by means of red hot clay placed over the
bottom and round the sides of a hole preparedfor that purpose. Over the hot clay they place
a thin layer of damp grass, upon which theylay the joint to be cooked, covering it over also
with damp grass, upon which more hot clayis placed, the whole is then carefully coveredwith sand. It is a very perfect method, and canbe made large enough to roast an ox, or small
enough to cook an opossum'. Similar methodsare still used by some tribes of Aboriginals,
e.g. at Yirrkala in N. Australia.
Brough Smyth (1878) stated that the LowerMurray natives 'cook their large game in ovensmade in the following manner—a hole is madein the earth and lined with stones; in it theymake a fierce fire, until the stones becomealmost red hot; the kangaroo, or what they
88 EDMUND D. GILL
intend to cook is then placed in the oven, on
the top of which some more hot stones are
placed, and the whole covered with earth; the
heat of the stones and the confined steam
together cook the meat' (Vol. 2, p. 298).
Beveridge's and Brough Smyth's descriptions
are important in that they explain that the ovens
were used for thick meat. In the lunette on the
E. side of Lake Victoria, numerous shell mid-
dens are found, and fossil extinct marsupials
are also found there, so some have inferred
that this proves the Aboriginals did not use
these marsupials for food. This in an incorrect
inference. It should be noted that the Abo-riginals employed two cooking media—the
small midden fire for cooking small pieces of
meat such as mussels, and the oven, the heat-
holding capacity of which permitted the cook-
ing of thick pieces of meat. Only once werebones found in a midden, and that was in the
one on the N. shore of Lake Victoria E. of the
old hotel site. This contained some bones of
the small marsupial Onchogalea jrenata. Thus,the composition of shell middens is not a guide
to what marsupials were used for food.
Unfortunately, no bones have been found in
ovens in the study area. It has been noted howthe bones of both modern and fossil animalsreadily decrepitate in this climate when they
are exposed at the surface. No bone implementswere found by us. Gallus (this Memoir) de-scribes two bone fish-hooks associated with a
skeleton. Brough Smyth (1878 vol. 1, p. 203)believed them to be absent, so perhaps their usewas prehistoric.
Ovens were used by the earliest known Abo-riginals, evidence of whose presence has beendiscovered at Lake Mungo, NE. of our studyarea (Bowler et al. 1970). Burnt clay has beendated there at 30,750 yr B.P. (Anonymous1972, Barbetti and McElhinny 1972). Duryand Langford-Smith (1970) recorded a fire-
place at Lake Yantara in NW. New SouthWales that dated 26,200 yr B.P. Judging by thephotograph, lumps of burnt sediment were pre-sent.
As these pieces of 'brick' could be usedagain and again, it might be assumed that whenrain fell, sediment on site was puddled andburnt. Ecologic analysis shows that in some
sites at least this can be proved untrue. Thus
ovens on dunes of well-sorted sand have cook-
ing stones made of relatively poorly-sorted
sediments from the nearby river.
Finally, it should be kept in mind that clay
balls were also used as sinkers (Hardy 1969,
p. 12).
c. The hearth. This third kind of fireplace
has ashes and charcoal only (with burnt ground
at times). No shells, bones or cooking stones
are present. The localized nature of the deposit
and its very close similarity to the middens andovens shows that it is of Aboriginal origin.
The fireplaces preserved are of course only
a fraction of those made. If ten groups of
Aboriginals in this area each made one fire aday for 30,000 years, this would amount to
about 11,000,000 fireplaces.
(d) Stone material. Being over a deep sedi-
mentary basin, the terrain is a stoneless land.
Then how did the Stone Age men manage for
stoneware? Their light industry (the smalltools ) was supplied from local pedogenicmaterials (opal, silcrete, carbonate), while their
heavy industry (millstones, mortars and such)were mostly imported from the areas where Pal-
aeozoic and Precambrian rocks from the mar-gin of the basin, viz. the Great Dividing Rangeto the E., the Barrier and other Ranges to N.,and the Mt. Lofty Ranges to the W. A peno-logical study to define origins (cf. Ambroseand Green 1972, Dixon et al 1968, Gloverand Cockbain 1971, Shotton 1968, Sievekinget al 1970) is needed.
Local rocks, even when unsuitable, weretried. Mr and Mrs H. Hansen collected a mill-
stone made of a relatively soft calcareous sand-stone, the source of which was probably PineCamp Station, where it outcrops at the side of adry lake. The best known local development ofpedogenic siliceus sandstone is at Moorna Sta-tion homestead where it was used for building.The paleosol that yielded it is now under water,as the river level is controlled in this reach ofthe river. This rock was utilized by the Abo-riginals.
Because all the local rocks are pedogenicmaterials of a limited range of types, it is easyto recognize the 'stranger rocks' brought in byhuman agency. It has already been mentioned
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 89
that the local common opal was used and not
the Mt. Gambier flint (Daley 1926), which wastraded up to the N. Mallee not far from the
Murray R. Only two pieces of flint were foundamong the many thousands of Aboriginal imple-
ments and flakes examined. Tertiary marinefossils were recognizable in them, and White-
head (this Memoir) reports on the application
of a new method in an endeavour to prove
their origin.
As the Aboriginals brought in by trade silici-
fied sandstone and quartzite for their large
implements, it might be expected that they
traded out other things such as opal. (Ontrading see Mulvaney 1967, p. 94 ff., Tindale
1968). However, Sir Robert Blackwood and I
as a check collected opal-like material frombetween Wentworth and Broken Hill; also at
Menindee and Mootwingee. Dr E. R. Segnit
examined these, but they all proved to be very
fine-grained silcrete and not opal. It is knownthat the Aboriginals traded reed stems fromthe Murray R. for spear shafts (Gutheridge
1907, Stone 1911).
On the lunette, a yellow Tertiary fossil
shark's tooth, Isurus hastcdis, was found. Thenearest formation from which this could comeis Miocene limestone that outcrops on the banksof the Murray 150 km to the W.
The range of Aboriginal implements foundin the area was a surprise. No pirris (Campbell
1960) or tula adzes (or a few doubtfully) werefound although these are the 'two basic pre-
historic implement types'. So says Mulvaney(1964) and added that they 'appear inexplic-
ably to observe State boundaries in their
occurrence' They were not found in the study
area nor in some places examined further N.,
so in this region the boundary is N. of the
State boundary. Mr T. Brown (pers. comm.)of Broken Hill has a good collection but they
came chiefly from N. of that town. Rare ground-
edge axes and cylcons have been reported, but
these uncharacteristic occurrences do not reflect
the local material culture, but perhaps strange
objects accepted as having magical power. Thefact is that the plentiful artefacts of the area
demonstrate the primitive core/flake culture,
such as characterizes the earliest known Abo-riginals.
The distribution of artefacts is instructive.
They occur in high concentrations beside LakeVictoria and at sites such as (a) The inland
limit of the lunette sand encountered on the
Talgarry pipeline track from Lake Victoria to
the dams near the homestead, (b) On the
same track SW. of the second dam from the
road (on the fencelinc). Both these sites are
in geomorphic lows with sand underlain by clay.
After heavy rain, pools of water last up to twomonths, and even when they have disappeared
water is available in the sand. Thus we have apicture of the Aboriginals as river people whomoved inland as pool-campers after the rains.
Some sites are more than 15 km from the
river/lake system.
The Aboriginal skeletons collected (Black-
wood and Simpson, Sandison, this Memoir)and those seen but not collected, indicate a
comparatively healthy people, as far as this
can be judged from their bones. Broad basedon the productive river/lake ecosystem, they
were apparently well fed (mussels, fish, cray-
fish, turtles, water-fowl, emus, marsupials,
plants). In addition, they were conservative,
adhering to the ancient core/flake industry, andallowing the new cultures of hafted tools andmicroliths to go by. They imported only whatwas necessary, relying on local materials. Theiroutside contacts were minimal, so limiting the
likelihood of infection from elsewhere.
The natives ground nardoo (Howitt 1908,Lees 1915) on their millstones, and used
hammerstones on mortars for a wide range of
jobs (McCarthy 1967). Canoes were used(Mitchell 1839, Beveridge 1865, Curr 1886)on the rivers and lakes (Mitchell 1839), and'canoe trees' still remain with the scars (nowwell overgrown at the edges) whence the barkwas removed. Scars left by bark removed byearly settlers for tables and such are some-times confused with Aboriginal canoe trees. Wetherefore experimented by removing squares of
bark from the same red gum with a steel toma-hawk and an Aboriginal stone axe of similar
size. The former cut sharp narrow deep cuts
with sides approximately parallel, while the
latter cut shallower burred cuts with curvedsides. We found that a sharp stick was also
effective.
90 EDMUND D. GILL
The reed spear-shafts of the river people
were too light for stone heads, so the tips were
made from mulga (Acacia). Mr J. Higgins (the
oldest resident) described how the Aboriginals
cut a groove round the mulga branch then cut
thick slivers towards the groove, thus obtaining
pieces of this very hard wood about the size
and shape of the quartzite flakes used elsewhere
for the same purpose. We did not know that
mulga occurred this far south, but Mr Higgins
pointed out two relict patches, one of over 10hectares and the other with only a few trees.
The literature describes pieces of stone used
for various types of magic by Aboriginals.
Some formless pieces of imported stone found
in the study area are presumed to have hadsome such use. For example, a flat piece of
flakey mica schist from Ned's Corner Station
was useless as an implement but its bright
golden flecks could well have suggested someinnate power to an Aboriginal.
McCarthy (1968) has described coastal andinland sequences of cultures with the Dividing
Range between as a barrier. Such barriers are
as much psychological (accepted boundaries)
as they are physical. Similarly, the MurrayRiver was a barrier. These barriers can beregarded as diffusion lines through which dif-
fusion vectors (usually southwards) operated
to bring about cultural change. The Murray R.constitutes a strip ecology surrounded by semi-
arid country, so it is a natural interface, andtribal boundaries were related to it. Tindale
(pers. comm.) states that the Maraura tribe
lived to the N. and the Ngintait tribe to the
S. The Maraura controlled the opal mines, andthis is probably why large lumps are often foundN. of the river but only small pieces in smaller
quantity S. of it. N. of the Murray R. werequartzitic millstones and mortars, pirris, tulas,
flake knives and picks, stone-tipped spears,
bull-roarers, cylcons, carved trees, copi grave
markers and so on. To the S. were basalt mor-tars, flint artefacts, wooden spears with hard-ened tips, and so on. However, in our study
area only the large flat millstones appear to
have the Murray R. as an actual limit of
distribution. They are common N. of the river,
but of negligible occurrence S. of it.
In review, one is amazed that in spite of the
changing ecology (climate, fauna, etc.) there
was a technological stability that lasted in this
area for some 30,000 years. In spite of this,
the Aboriginals of the area readily took to
European meats, tomahawks, knives, blankets,
and so on (Mitchell 1839).
Economic Considerations
The soil is this country's greatest wealth, so
it needs conservation. Some of the world's
finest wools are grown in this semi-arid country.
Water is the limiting factor. In the dry summertime, especially in drought, the terrain appears
harsh, but when the heavy rains fall, a miracle
of regeneration is re-enacted. The perennials
spring up with amazing rapidity and in six to
eight weeks mature and seed. As the irrigation
areas prove, this land is most productive whenwater is available.
South of the Murray R., the dominance of
sandy terrain results in losses of up to 90%of the water passed through the feeder channels.
N. of the river the Blanchetown Clay (below
the veneer of sand) permits the excavation of
earth tanks to preserve run-off water. TheDarling R. is the local Nile in that it begins
in a humid zone then flows 2,500 km and morethrough dry country. The clay floor and bankspreserve its waters which are not supplementedby tributary streams.
On this semi-arid terrain there is one sheepto 2-5 hectares (6 acres) or more. The large
woolshed of the Lake Victoria station has pro-
cessed up to 98,700 sheep at one shearing
(32 stands). Some cattle are raised along the
rivers and lake shores. Cattle require fresher
water than sheep which will drink water withup to 7,000 ppm (weight), i.e. about one fifth
the salinity of the sea.
The water table in this area is very low.Piezometer bores in connection with theChowilla Dam (data kindly made available byGutteridge, Haskins and Davey) proved oxida-tion from the general land surface about 37 mabove the river up to 12 m below river level,
and up to 8 5 below water level in August1968. The pH of the groundwater varied from7 1-8-3. Sodium and chloride constituted77- 1% of total salts. Water from all the bores(traverse lines across the Murray R. from the
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 91
S.A. border to Mildura) had much the sameratio of salts.
The salinity of groundwaters and river
waters has been a major problem. During the
drought of 1967 the River Murray Commissioncontracted Gutteridge, Haskins and Davey in
association with Hunting Technical Services
Ltd. to study this matter and in 1970 three
volumes were published, viz. 1. The Report,
2. Maps, and 3. A Summary. These volumesdefine the amount of salt involved, and its
effects under the varying conditions obtaining
in that area from time to time. The Reportindicates the need of further research and the
importance of adequate management.Salt is harvested at Lake Tyrrell. Papers on
salinity in the Murray region include Anony-mous 1968, Crabb 1968, Hutton 1958, 1969and references, Livermore 1968, McLaughlin1966, Penman 1969, Williams 1970. Papers onunderground water include Barnes 1951, Gib-bons et al. 1972, Lawrence 1966, Macumber1968, O'Driscoll 1966 and references.
Gypsum has been worked commercially in
the study area. Timber is gained, chiefly river
red gum. Commercial fishermen operate along
the Murray R.
Beds of dolomite are reported for the first
time in this area, but at present there is nocommercial use for it.
Conclusion
This virtually unknown piece of country hasproved to have rich geologic, archaeologic andbiologic histories. The present reconnaisance,
reported here inadquately under pressure of
time, is but an indication of what has yet to
be discovered. There are so many obvious
worthwhile research projects and I am so often
asked for suggestions, that I append a list.
APPENDIX 1
List of MapsMilitary maps 1" = 4 mi. Renmark, Chowilla, Mil-
dura, Anabranch.Fire maps Wentworth.Photo maps Lindsay, Wentworth, Mil-
dura.
E.W.S. 1-6.
Air photos Lindsay, Wentworth,Lake Victoria.
APPENDIX 2
Research Projects
1. Relationship of stream direction to tectonics.
2. The system of exceptionally fine-grained fluvia-
tile, lacustrine, beach and dune sediments..3. Age and evolution of Lake Victoria.4. The Boy Creek/Wallpolla Creek anabranch sys-
tem.5. The remarkably extensive meander belts.
6. The age of the major oxbows as shown by radio-carbon dating of their basal sediments.
7. The origin of mottled profiles below present baselevel (e.g. Moorna East oxbow lake).
8. Analysis of meander lengths of the present andthe past.
9. The differences in lithology and pedology N. andS. of the Murray River in the study area.
10. Origin of depressions (such as hold salt lakes)in the flat plain.
11. Stratigraphy of Salt Lake on Talgarry Stationto follow its origin and history, dating by radio-carbon.
12. Relationships of E-W. longitudinal dunes (Woori-nen Formation) with other formations (e.g.
Rufus Formation), and with the various levelsof Murray River incision.
13. Dunefield history, and the nature of the de-pression of dune forms in each stability phase,marked by near horizontal soils.
14. Differences between Mallee and Lake VictoriaDunefields, and the reasons therefor.
15. The relationship between Lowan Sands and theMallee E-W. Dunefield.
16. The relationship of the present Lake Victoria tothe basin in which it lies.
17. The gulches in the lunette on the E. side ofLake Victoria (which present extensive strati-graphic, pedologic, sedimentologic, archaeologicand palaeontologic data).
18. The facies of the Lake Bungunnia deposits, andtheir palaeogeography.
19. Survey the Triple Swamp cliffs, Fisherman's Cliff,and the cliffs of Moorna E. oxbow lake (whichare nearly continuous) to provide a detailedsection many kilometres long that would elucidatethe lateral variation and interdigitation of for-mations in this area.
20. The Rufus Formation, its character and extent.It offers the opportunity of discovering any sig-nificant change in the Murray R. tract and theLake Victoria basin since the ? Middle Pleis-tocene. Also its relationship to ancestral andprior streams.
21. The Monoman Formation, its distribution up-stream, and its relationship with the ancestraland prior streams; also with change of sealevel.
22. Systematic C14 and other dating of paleosols inthe Lake Victoria lunette and the E-W. dunes(Woorinen Formation) to define the periods ofstability (soil-forming) and instability (dune-building) on this terrain. This would constitute avaluable contribution to palaeoclimatology.
23. Along with 22 or as a separate project to carryout grids of grain size analyses to discover (a)the degree of consistency in the sediments and(b) whether certain types (e.g. bimodal) can be
92 EDMUND D. GILL
traced through the system. Study the mineral-
ogy of clays and other minerals of the dunesand associated sediments in an effort to find
out which are locally derived and which areadditives from afar.
24. Take undisturbed cores from the surface to the
Bookpurnong Formation through the type MoornaFormation section to study the sediments andfossils in detail. Check if the Parilla Sand occursbelow the surface, and the nature of the change-over from marine to riverine sedimentation.
25. Definition of the incoming of the present semi-aridity by detailed investigation of the strati-
graphy, palaeontology (including palynology),mineralogy and palaeoecology of the horizonconcerned.
26. Investigate the bones in the Nulla Nulla SandMember which develop a cheese-like consistencyin very wet weather, and may be deformed whilein this plastic condition.
27. Describe the ecosystems of the area, and studytheir evolution.
28. Petrol ogic examination of the stones used by theAboriginals to define their origins.
References
Ahnert, F., 1970. Functional relationships betweendenudation relief uplift in large mid-latitudedrainage basins. Am. J. Sci. 268: 243-263.
Ambrose, W. R., and Green, R. C, 1972. Firstmillenium B.C. transport of obsidian from NewBritain to the Solomon Islands. Nature 273: 31.
Andrews, E. C, 1910. Geographical unity of EasternAustralia in Late and Post-Tertiary Times, withapplications to biological problems. J. R. Soc.N.S.W. 44: 420-480.
Anonymous, 1857. Specimens exhibited—Rocks andfossils of a highly interesting character from thebasin of the Murray. Trans. Phil. Inst. Vict. 1:
XXV.Anonymous, 1968. Salinity in the River Murray.
Nature 220: 537-538.Anonymous, 1972. The Mungo reversal. Search 3
(3): 51.Axelrod, E. I., and Bailey, H. P., 1969. Paleotem-
perature analysis of Tertiary floras. Palaeogeogr. t
Palaeoclimat., Palaeoecol. 6: 163-195.Aziz-ur-rahman, and McDougall, I., 1972. Potas-
sium-argon ages on the Newer Volcanics ofVictoria. Proc. R. Soc. Vict. 85: 61-69.
Barbetti, M., and McElhinny, M., 1972. Evidenceof a geomagnetic excursion 30,000 yr B.P. Nature.239: 327-330.
Barnard, F. G. A., 1904. Some early botanical ex-plorations in Victoria. Vict. Naturalist. 21: 17-28(with map).
Barnes, T. A., 1951. Underground water survey ofportion of the Murray Basin. Geol. Surv. S.Aust. Bull. 25.
Bettenay, £•„ 1962. The salt lake systems and theirassociated aeolian features in the semi-arid re-gions of Western Australia. J. Soil Sci. 13-10-17.
Beveridge, P., 1865. A few notes on the dialects,habits, customs and mythology of the LowerMurray aborigines. Trans. Proc. R. Soc Vict6: 14-24.
Blackburn, G., 1971. Fossil podocarp root from
Telangatuk East, Victoria. Vict. Naturalist 88:
56-57.
Blandowski, W., 1857. Recent discoveries in natu-
ral history in the Lower Murray. Trans. Phil.
Inst. Vict. 2: 124-137.
Bonwick, J., 1855. Geography of Australia and NewZealand. Melbourne, 3rd ed.
Bonython, C. W., 1955. Lake Eyre, South Australia
—The great flooding of 1949-50. R. geogr. Soc.
A'asiat S.A. Branch, Adelaide, pp. 7-9, 27-36,
37-56, 63-68, 69-70.
Bowler, J. M., 1967. Quaternary chronology of TheGoulburn Valley sediments and their correlation
in southeastern Australia. J. geol. Soc. Aust. 14:
287-292.Bowler, J. M., and Harford, L. B., 1963. Geomor-
phic sequence of the riverine plain near Echuca.Austr. J. Sci. 26: 88.
, 1966. Quaternary tectonics and the evolu-tion of the Riverine Plain near Echuca, Victoria.
/. geol. Soc. Aust. 13:
-, Jones, R., Allen, H., and Thorne, A. G.t
1970. Pleistocene human remains from Aus-tralia: A living site and human cremation fromLake Mungo, Western New South Wales. WorldArchaeology, May 1970.
and Macumber, P. G., 1968. The riverineplain in Northern Victoria. Regional Guide toVictorian Geology (pp. 133-144). Ed. J. Mc-Andrew and M. A. Marsden, Melbourne.
Boyland, D. E., 1971. Ecological and floristic studiesin the Simpson Desert National Park, S.W.Queensland. Proc. R. Soc. Qd. 82: 1-16.
Brewer, R., Crook, K. A. W., and Speight, J. G. f
1970. Proposal for soil-stratigraphic units inThe Australian Stratigraphic Code. /. geol. Soc.Aust. 17: 103-111.
Bride, T. F. (Ed.) 1898. Letters from VictorianPioneers. Melbourne.
Brough Smyth, R., 1878. The Aborigines of Vic-toria (2 vols.) Melbourne.
Browne, W. R., 1934. Some peculiarities in thedrainage-systems of the Australian continent.Aust. Geogr. 2 (4): 13-19.
Butler, B. E., 1950. A theory of prior streams asa causal factor in the distribution of soils inthe riverine plain of S.E. Australia. Aust. J.agric. Res. 1: 231-252.
, 1958. Depositional systems of the riverineplain of SE. Australia in relation to soils. SoilPuhl. CSIRO Aust. 10.
, 1960. Riverine deposition during aridphases. Aust. J. Sci. 22: 451-452.
, 1961. Ground surfaces and the history ofthe riverine plain. Aust. J. Sci. 24: 39-40
Cameron, W. J. (Ed.) 1966. The History of Bourke,being papers prepared by members of TheBourke Historical Society 1964-6. 148 pp. (dupli-cated).
Campbell, E. M., 1968. Lunettes in southern SouthAustralia. Trans. R. Soc. S. Austr. 92: 85-109.
Campbell, T. D., 1960. The pirri—an interestingAustralian aboriginal implement. Rec. S. AustMus. 13: 509-524.
Chapman, F.. and Grayson, H. J., 1903. On 'RedRain*, with special reference to its occurrencein Victoria. With a note on Melbourne dustVict. Naturalist 20: 17-32.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 93
Chico, R. J., 1968. Playa. Encyclopedia of Geomor-pholoy (Ed. R. W. Fairbridge), pp. 865-871.New York.
Chorley, R. J., and Beckingsale, R. P., 1968. Baselevel. Encyclopedia of Geomorphology (Ed.
R. W. Fairbridge), pp. 58-60. New York.Churchward, H. M., 1961. Soil studies at Swan Hill,
Victoria, Australia.
1. Soil layering. /. Soil Sc. 12: 73-86.
, 1963 a. Soil studies at Swan Hill, Victoria,
Australia.
2. Dune moulding and parna formation. Aust.
J. Soil Res. 1: 103-116.
-, 1963b. Soil studies at Swan Hill, Victoria,
Australia.
3. Some aspects of soil development on aeolian
materials. Aust. J. Soil Res. 1: 117-128.
-, 1963c. Soil studies at Swan Hill, Victoria,
Australia.
4. Ground surface history and its expression in
the array of soils. Aust. J. Soil Res. 1: 242-
255.
Crabb, P., 1968. Some aspects of salinity in the South
Australian Upper Murray. Aust. Geogr. vl: 170-
175 (inch Reply by J. F. Livermore).
Crocker, R. L., 1946. Post-Miocene climatic andgeologic history and its significance in relation
to the genesis of the major soil types of South
Australia. CS1RO Aust. Bull. 193.
Crook, Keith A. W., 1967. Tectonics, climate andsedimentation. 7th Int. Congr. Sediment., Read-
ing, England.Curr, E. M., 1886. The Australian Race. Melbourne.
4 vols.
Daley, C, 1926. The use of flint among Australian
Aborigines. Rept. Aust. Assoc. Adv. Sci. 18:
519-520.Davis, W. M., 1954. Geographical Essays. New York,
Dover 777 pp.
Dawson, J. W., 1970. Rain forests and Gondwanaland,Tuatara 18 (2): 94-5.
, and Sneddon, B. V., 1969. The New Zea-
land rain forest; a comparison with tropical rain
forest. Pacif. Sci. 23: 131-147.
Devine, P. E., and Power, S. B., 1970. Surat Basin,
Australia—subsurface stratigraphy, history and
petroleum. Bull. Amer. Assoc. Petrol. Geol. 54:
2410-2437.
Dixon, J. E., Cann, J. R., and Renfrew, C, 1968.
Obsidian and the origins of trade. Scient. Ameri-
can 218 (3): 38-46.
Duigan, Suzanne L., 1966. The nature and relation-
ships of the Tertiary brown coal flora of the
Yallourn area in Victoria, Australia. The Palaeo-
botanist 14: 1-3.
Dury, G. H., 1960. Misfit streams: problems in in-
terpretation, discharge and distribution. Geogr.
Rev. 50: 219-242.
, 1963. Prior stream deposition. Aust. /.
&A 25: 315-317.
-, 1965. Theoretical implications of underfit
streams. U.S. geol. Surv. prof. Pap. 452-C, 43 pp.
-, and Langford-Smith, T., 1970. A Pleisto-
cene Aboriginal camp fire from Lake Yantara,
NW. NSW. Search 1 (2): 73.
Edwards, R., 1967. A survey of Aboriginal relics in
the Chowilla area by the South Australian Mus-
seum. Aust. Inst. Abor. Stud. Newsletter 2(6): 42-44.
Edwards, A. B., and Baker, G., 1951. Some occur-
rences of supergene iron sulphides in relation
to their environment of deposition. /. sed. Petr.
21: 34-36.Everard, George, n.d. Pioneering days. 34 pages
(duplicated).
Fenner, C, 1934. The Murray River Basin. Geogr.Rev. 79.
Firman, J. B., 1963. Quaternary geological events
near Swan Reach in the Murray Basin, SouthAustralia. Quart, geol. notes (Geol. Surv. S.A.)
5.
, 1965. Late Cainozoic lacustrine deposits in
the Murray Basin, S.A., Quart, geol. Notes (Geol-
Surv. S.A.) 16: 1-2.
1966. Stratigraphy of the Chowilla area in
the Murray Basin. Quart, geol. Notes (Geol.
Surv. S.A.) 20: 3-7.
-, 1967a. Late Cainozoic stratigraphic units in
South Australia. Quart, geol. Notes (Geol. Surv.
S.A.) 22: 4-8.
-, 1967b. Stratigraphy of late Cainozoic de-
posits in South Australia. Trans. R. Soc. S. Aust.
91: 165-178.-, 1969. Stratigraphic analysis of soils near
Adelaide, South Australia. Trans. R. Soc. S. Aust.
93: 39-54.
-, 1970. Structural lineaments in the MurrayBasin of South Australia. Quat. geol. Notes(Geol. Surv. S.A.) 35: 1-3.
1971a. Regional stratigraphy of surficial
deposits in the Murray Basin and Gambier Em-bayment. S. Aust. Dept. Mines Rept. Bk. 71/1.
-, 1971b. Riverine and swamp deposits in the
Murray Tract, S. Australia. Quat. Geol. Notes(Geol. Surv. S.A.) 40: 1-4.
Folk, R. L., 1969. Grain shape and size in the SimpsonDesert, NT., Australia. Geol. Soc. Amer. Ab-stracts with Programs for 1969 Pt. 7, pp. 68-69-
Folk, R., 1971. Genesis of longitudinal and oghurddunes elucidated by rolling upon grease. Geol.
Soc. Amer. Bull 82: 3461-3468.
Fuller, J. G. C. M., 1971. The geological attitude.
Bull. Am. Assoc. Petrol. Geol 55: 1927-1938.
Gibbons, G. S., Griffin, R. J., and Staude, W. J.,
1972. A review of ground-water in the MurrayBasin, N.S.W. Bull, geol Surv. N.S.W. 19.
Gill, E. D., 1953. Geological evidence in WesternVictoria relative to the antiquity of the Australian
Aborigines. Mem. nat. Mus. Melb. 18: 25-92.
, 1957. The stratigraphical occurrence andpalaeoecology of some Australian Tertiary mar-supials. Mem. nat. Mus. Melb. 21: 135-203.
-, 1961. The climates of Gondawanaland in
Cainozoic time. Ch. 14 of Descriptive Palaeo-climatology, Ed. Nairn.
-, 1961. Cainozoic climates of Australia. Ann.New York Acad. Set 95 (1): 461-464.
-, 1964. Rocks contiguous with the basaltic
cuirass of Western Victoria. Proc. R. Soc. Vict.
77: 331-355.-, 1965 . Quaternary geology, radiocarbon
datings and the age of australites. Geol. Soc.Amer. Spec. Pap. 84: 415-432.
-, 1966. Geochronology of Victorian Malleeridges. Proc. R. Soc. Vict. 79: 555-559.
94 EDMUND D. GILL
-, 1967b. Australian Aboriginal remains5540 years old from Mitiamo, Victoria, Australia.
Proc. R. Soc. Vict. 80; 289-293.- 1969. Some aspects of prehistory in Vic-
toria. Vic. hist. Mag. 40: 173-189.
1970. Rivers of history, A.B.C. Sydney,67 pp.—, 1971a. Laterite chronology. Search 2: 32.
- 1971b. Applications of radiocarbon datingin Victoria, Australia (Research Medal Lecture).Proc, R. Soc. Vict. 84: 71-85.
-, 1972a. Tectonics and landforms of theNew Guinea region: review of a symposium.Pacif. Geol. 4, 23-26.
-, 1972b. The Pliocene-Pleistocene boundaryin Australia: a review. INQUA International Col-loquium on the Problem 'The Boundary betweenNeogene and Quaternary' Collection of Papers1: 63-71. Moscow.
-, 1972c. Eruption date of Tower Hill vol-
cano, Western Victoria, Australia. Vict. Natu-ralist 89: 188-192.
1972d. Palaeoclimatology and dinosaurs
in South-east Australia. Search 3: 444-446.
Gill, E. D., and Bethune, F. N., 1976. Radiocarbondates from Narrandera, N.S.W., Australia. Vict.
Naturalist 84: 125-127., and Hopley, D., 1972. Holocene sea levels
in Eastern Australia: Discussion. Marine Geol.
12: 223-242.Glasby, G. P., 1971. The influence of aeolian trans-
port of dust particles on marine sedimentationin the south-west Pacific. /. R. Soc. N.Z. 1:
285-300.Gloe, C. S., 1947. The underground water resources
of Victoria. S.R.W.S.C. Melbourne.Glover, J. E., and Cockbain, A. E., 1971. Trans-
ported Aboriginal artefact material, Perth Basin,Western Australia. Nature 234: 545-6.
Guthridoe, J. T., 1907. The Stone Age & Aboriginesof the Lancefield District. Castlemaine, 11 pp.
Hansen, R. O., and Stout, P. R., 1968. Isotopic dis-
tributions of uranium and thorium in soils. Soil
Set. 105: 44-50.
Hardy, A. D., 1914. The Mallee: Ouyen to Pinnaroo.Vict. Naturalist 30: 148-167.
Harcy, Bobbie, 1969. West of the Darling. Milton,Queensland.
Harris, W. J., 1939. The physiography of the EchucaDistrict. Proc. R. Soc. Vict. 51: 45-60.
Hawdon, J. 1852. The Journal of a journey fromN.S.W. to Adelaide (performed in 1838). Mel-bourne.
Heathcote, R. L., 1969. Drought in Australia: aproblem of perception. Geogr. Rev. 59: 175-194.
Hills, E. S., 1939. The physiography of NW.Victoria. Proc. R. Soc. Vict. 51: 293-320.
, 1940. The lunette, a new land form ofaeolian origin. Austr. Geogr. 3 (7): 1-7.
1956a. A contribution to the morphotec-tonics of Australia. /. geol. Soc. Aust. 3: 1-15.
-, 1956b. The tectonic style of Australia.Geotckt. Symp. Ehr. Hans Stifle, pp. 336-346.
1961. Morphotectonics and the geomorpho-logical sciences with special reference to Aus-tralia. Q JL geol. Soc. Lond. 117: 77-89.
1966. (Ed.) et al., Arid lands—a geo-graphical appraisal. UNESCO, p. 461.
Howitt, A. W., 1908. Personal reminiscences of
Central Australia and the Burke and Wills Ex-pedition. Rept. (1 1th Mtg.) Aust. Assoc. Adv.Sci. 11; 1-43.
Hutton, J. T., 1958. The chemistry of rainwater withparticular reference to conditions in SE. Aus-tralia. UNESCO Arid Zone Res. 11: 217-221.
, 1969. The redistribution of the moresoluble chemical elements associated with soils
as indicated by analysis of rainwater, soils andplants. Trans. 9th Internal. Congs. Soil Sci.
4: 313-322.
Jackson, Colonel, 1834. Hints on the subject of
geographical arrangement and nomenclature. /. R.geogr. Soc. 4: 72 ff.
Jennings, J. N., 1968. A revised map of the desert
dunes of Austr. Aust. Geogr. 10: 408-409.
Johns, M. W., and Lawrence, C. R. t 1964. Aspectsof the geological structure of the Murray Basinin NW. Victoria. Geol. Surv. Vict. UndergroundWater Invest. Rep. 10.
Johns, R. K., 1962. Investigation of Lake Eyre. GeoLSurv. Aust. Rept. Invest. 24.
Jones, J. G., 1971. Australia's Caenozoic drift. Nature230: 237-239.
Jutson, J. T., 1934. The physiography of WesternAustralia. Bull. geol. Surv. W. Austr.
Kenyon, A. S., 1942. The story of the Murray River.
Vict. Naturalist 58: 193-6; 59: 16-19.
Kino, D., 1956. The Quaternary stratigraphic record
at Lake Eyre North and the evolution of existing
topographic forms. Trans. R. Soc. Aust. 79: 93-
103.
Kisutsina, G. I., and Cherdyntsev, V. V., 1967.
Metodika opredeleniya absolyutnogo vozrasta
pleystotsenovykh formatsiy po neravnovesnomuuranu (Age determinations of Pleistocene for-
mations by the disequilibrium of uranium)
.
Akad, Nauk SSSR Kom. Opredeleniyu Absolyut.Vozrasta Geol. Formatsiy Byull. 8: 182-185.
Krumbein, W. C, and Sloss, L. L., 1963. Strati-
graphy and sedimentation. San Francisco.
Kulp, J. L-, 1961. Geologic time scale. Science 133:1105-1114.
Lake, J. S., 1967. Principal fishes of the Murray-Darling river system. Australian Inland Watersand their Fauna (Ed. A. H. Weatherly) A.N.U.,Canberra, pp. 192-213.
Langbein, W. B., and Schumm, S. A., 1958. Yieldof sediment in relation to mean annual precipi-
tation. Am. Geophys. Union Trans. 39: 1076-1084.
Langford-Smith, T., 1960. The dead river systemsof the Murrumbidgee. Geogrl. Rev. 50: 368-389.
, 1962. Riverine plains geochronology. Aust.
J. Sci. 25: 96-97.Lawrence, C. R., 1966. Cainozoic stratigraphy and
structure of the Mallee Region, Victoria. Proc.R. Soc. Vict. 79: 517-553.
Lawrence, R., 1968. Aboriginal habitat and economy.A.N.U. Dept. Geogr. Occas. Paper 6.
Lees, E. H., 1915. What is Nardoo? Vict. Naturalist
31: 133-135.Leichardt, L., 1847. Journal of an overland expedi-
tion in Australia etc. London.Livermore, J. F., 1968. Some aspects of salinity in
the South Australian Upper Murray. A ustn.
Geogr. 10 (6): 520-522.Lowry, D. C, 1970. Geology of the Western Aus-
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 95
tralian part of the Eucla Basin. Bull, geol. Surv.W. Aust. 122.
Ludbrook, N. H., 1961. Stratigraphy of the MurrayBasin in South Australia. Geol. Surv. S. Aust.Bull. 36.
, 1969. Tertiary foraminiferal zones in SouthAustralia. Proc. First Internat. Conj. Plank.Microfossils 2: 366-375.
-, et at, 1958. The Murray Basin in SouthAustralia. /. Geol. Soc. Aust. 5 (2): 102-114.
Mabbutt, J. A., 1968. Aeolian landforms in CentralAustralia. Aust. Geogr. Stud. 6: 139-150.
-, and Sullivan, M. E., 1968. The formationof longitudinal dunes: evidence from the Simp-son Desert. Austr. Geogr. 10 (6): 483-487.
McCarthy, F. D.. 1967. Australian Aboriginal StoneImplements. Australian Museum, Sydney.
, 1968. Recent developments and problemsin the prehistory of Australia. Paideuma, Mitteil.
Kulturkunda Bd. 14: 1-16.
McLaughlin, R. J. W., 1966. Geochemical concentra-tion under saline conditions. Proc. R. Soc. Vict.
79: 569-577.Macumber, P. G., 1968. Interrelationship between
physiography, hydrology, sedimentation andsalinisation of the Loddon River plains, Australia.
/. Hydro 1 . 7: 39-57.
, 1969. The inland limits of the Murravianmarine transgression in Victoria. Aust. J. Sci.
32: 165-166.1970. Lunette initiation in the Kerang
District Min. geol J. (Vict.) 6 (6): 16-17.
Mawson, D., 1950. Occurrence of water in LakeEyre, South Australia. Nature 166: 667.
Mitchell, T. L., 1839. Three expeditions into the
interior of Eastern Australia; with descriptions
of the recently explored region of Australia
Felix, and of the present colony of N.S.W. 2vols. London.
Morton, W. Lockhart, 1861. Remarks on the phy-sical geography, climate etc. of the regions lying
between the rivers Lachlan and Darling. Trans.
R. Soc. Vict. 5: 128-140.Mulvaney, D. J., 1964. Prehistory of the basalt
plains. Proc. R. Soc. Vict. 77: 427-432.
, 1969. Prehistory of Australia. London.Norris, R. M., 1969. Dune reddening and time.
/. Sed. Petr. 39: 7-11.
O'Driscoll, E. P. D., 1960. The hydrology of the
Murray Basin Province in S. Australia. Geol.Surv. S. Austr. Bull. 35.
Pels, S., 1964a. Quaternary sedimentation by prior
streams on the riverine plain, SW. of Griffith,
N.S.W. /. Proc. Roy. Soc. N.S.W. 97: 107-115.
, 1964b. The present and ancestral MurrayRiver system. Aust. geogr. Studies 2: 111-119.
- 1966. Late Quaternary chronology of the
riverine plain of SE. Australia. J. geol. Soc.
Aust. 13: 27-40.
1969. The Murray Basin. /. geol. Soc. Aust.
16 (1): 499-511.Penman, F., 1969. Soil and salinity factors in irri-
gation and drainage of Mallee lands. Trans. 9th
Internat. Congr. Soil Sci. 1: 415-423.
Roberts, G. T., 1968. Features of the BlanchetownClay in the Waikerie District. Quart, geol Notes(Geol. Surv. S.A.) 28: 4-6.
Shotton, F. W., 1967. The problems and contribu-tions of methods of absolute dating within thePleistocene period. Q. JL geol. Soc. Lond. 122:357-383.
, 1968. Prehistoric man's use of stone in
Britain. Proc. geol. Assoc. 79 (4): 477-491.Sieveking, G. de G., Craddock, P. T., Hughes, M. J.,
Bush, P., and Ferguson, J., 1970. Characteriza-tion of Prehistoric flint mine products. Nature228: 251-254.
Smith, A. G., and Hallam, A., 1970. The fit of thesouthern continents. Nature 225: 139-144.
Stamp, L. D., (Ed.), 1961. A glossary of geographicalterms. London.
Stannard, M., 1962. Prior stream deposition. Aust.J. Sci. 24: 324-325.
Stone, A. C, 1911. The Aborigines of Lake Boga,Victoria. Proc. R. Soc. Vict. 23: 433-468.
Sturt, C., 1849. Narrative of an expedition into
Central Australia, etc. London.Tate, R., 1885. Notes on the physical and geological
features of the basin of the Lower MurrayRiver. Trans. R. Soc. S. Aust. 7: 24-46.
, 1899. On some older Tertiary fossils ofuncertain age from the Murray Desert. Trans.R. Soc. S. Aust. 23: 102-111.
Taylor, G., 1914. The physical and general geo-graphy of Australia. Brit. Ass. Adv. Sci. Fed.Handb. pp. 86-121.
Tedford, R. H. 5 1967. The fossil macropidae fromLake Menindee. New South Wales. Univ. Calif.Publ, Geol. Sci. 156 pp.
Thomas, D. E., 1952. Geology, physiography andmineral resources. Victoria Resources SurveyMallee Region, pp. 1-2.
Thomas, M., 1971. Savanna lands between desert andforest. Geogr. Mag. 44: 185-189.
Tindale, N. B., 1968. Nomenclature of archaeologi-cal cultures and associated implements in Aus-tralia. Rec. S. Aust. Mus. 15: 615-640.
Turnbull, W. D., and Lundelius, E. L-, 1970. TheHamilton Fauna. A late Pliocene mammalianfauna from the Grange Burn, Victoria, Aust.Fieldiana: Geology Vol. 19.
Twenhofel, W. H., and Tyler, S. A., 1940. Methodsof study of sediments. New York.
TwroALE, C. R., 1972. Landform development in theLake Eyre region, Australia. Geogr. Rev. 62:40-70.
Veevers, J. L, Jones, J. G., and Talent, J. A., 1971.Indo-Australian stratigraphy and the configurationand dispersal of Gondwanaland. Nature 229:383-388.
Wakefield,, N. A. 1972. Palaeoecology of fossil
mammal assemblages from some Australiancaves. Proc. R. Soc. Vict. 85: 1-26.
Waldman, M., 197 1 . Fish from the freshwaterLower Cretaceous of Victoria, Australia, withcomment on the palaeo-environment. Spec. Pap.Palaeont. No. 9.
Walker, P. H., and Costin, A. B., 1971. Atmos-pheric dust accession in south-eastern Australia.Aust. J. Soil Res. 9: 1-5.
Wellman, P., McElhinney, M. W., and McDou-gall, I., 1969. On the polar-wander path forAustralia during the Cenozoic. Geophys. J. R.astr. Soc. 18: 371-395.
96 EDMUND D. GILL
Williams, W. D., 1970. Salt lake ecosystems. Aust.Sot: Limnol. Bull 3: 18-19.
Wilson, A. T., and Hendy, C. H., 1971. Past windstrength from isotope studies. Nature 234: 344-345.
Wopfner, H., and Twidale, C. R., 1967. Geomor-phological history of the Lake Eyre Basin. Land-form studies from Australia and New Guinea(Ed. J. N. Jennings and J. A. Mabbutt), pp.119-143. Canberra.
Zenger, D. H., 1972. Significance of supratidal dolo-mitization in the geologic record. Bull, geol. Soc.Am. 83: 1-12.
Explanation of Plates 1-11
Plate 1
Fig. 1—Murray River at Kulcurna homestead, CalLai, N.S.W. Low declivity and extensivefloodplain are characteristic of the rivertract. Eucalyptus camaldulensis lines theriver, while behind is Eucalyptus largi-
florens.
Fig. 2—Darling River below the bridge at Went-worth, N.S.W. Tucker Creek enters on thefar side. Telcphoto view (which puts theEucalyptus camaldulensis in the foregrounda little out of focus). Pelicans swim onthe placid waters. To the right of the photo,the river joins the Murray.
Plate 2
Fig. I—Bank of Murray River tract at Paringa, S.
Australia. Rilled Blanchetown Clay withParilla Sand below. Between is a band ofyellow silcrcte (Chowilla Sand) causing theplatforms at mid-cliff level. The silcrete is
the Karoonda Pedoderm, marking the Ka-roonda Terrain.
Fig. 2—Standard stratigraphic section at the pro-posed Chowilla Dam site (Fig. 4). Thesuccession is the same as at Paringa.
Plate 3
Floodplain of the River Murray, consisting ofCoonambidgal Formation sediments. N. is to thetop of the photo. The Murray R. appears to the SW.and Salt Creek (Kulcurna Stn.) to the E. withoxbows, anabranches, meander scrolls, and such. Onthe N. side of the oxbow in the NE. corner is TarcenaHomestead, now a pari of Kulcurna Stn. Last century,Cheltenhamian (late Miocene) marine fossils wererecovered from a well on Tareena Station. Air PhotoCAC 37-5004 Lake Victoria Run 5, 21 Feb. 1954.Crown copyright, published by permission of theDirector, Division of National Mapping, Dept.National Development, Canberra, A.C.T.
Plate 4
Air photo of Bernbce Station, NW. Victoria, Lind-say Run 2 1363/5245, reproduced by courtesy ofthe Under Secretary of Lands, New South Wales. N.is to the top of the photo. To the N. is the MurrayRiver, and S. of it is the anabranch called the LindsayR. Between is Lindsay Island. The prominent N-S.channel border dune at the W. end of the Island is
that shown in Fig. 11. The arrow marks the site of
the auger hole in the E-W. dune shown in Fig. 10.
Lindsay Island consists of Coonambidgal Formationsediments, while the SW. sector of the photo is
occupied by E-W. dunes (Woorinen Formation). Theauger hole is in an area of penetration of the E-W.dunes by riverine action. For general relationships
see Fig. 4.
Plate 5
Fig. 1—Lake Victoria, now used for water storage.
Photo shows lake at lowest level, and the
swash mark its top level. The dead trees markthe old water's edge. They now form nesting
places for birds. The sediment is fine sand.
Fig. 2—Grayish green Blanchetown Clay outcropshere above Sharp Point, and is surmountedby the rilled red soil profile at plateau level.
The white band at the base of this soil is
not carbonate, but a layer of white sandbelonging to the Blanchetown Clay forma-tion. Due to the semi-arid climate, the out-
crop is not overgrown. See cover picture.
Plate 6
Fig. 1—Murray River cliffs at the Kulcurna Stationhomestead. The lower part of the cliff is
Parilla Sand and the top part Chowilla Sand.The surface is the exhumed Karoonda Ter-rain. The building at the end of the cliff is
the remnants of the wharf whence wool wasloaded on to river boats in the early days.
Fig. 2—Bone Gulch, Moorna Station, W. of Went-worth, N.S.W. (Fig. 22). The wall of thegulch is mostly Blanchetown Clay, whichincludes layers of ostracods and some bones.The main ossiferous deposit is the ChowillaSand below. The fossils are mostly worn andare concentrated in the top of the formation,suggesting a natural sieving action.
Plate 7
Fig. 1—Gulch in Murray River cliffs below NampooStation Homestead, E. of Cal Lai, W. ofWentworth, N.S.W. The pine tree on theskyline is the position of the homestead. SeeFig. 4. The entire section consists of Blanche-town Clay, which is unvegetated in this semi-arid climate.
Fig. 2—Base of the section shown in Fig. 1. Underthe clay is a thin bed of Chowilla Sand andbelow that a formation of clayey sand whichshould probably be referred to the MoornaFormation. See Fig. 21.
Tig. 3*—Soil at plateau level, developed in Blanche-town Clay.
Fig, 4—Large rosettes of gypsum in BlanchetownClay in gulch in W. bank of Six Mile Cr.,
Neilpo Stn., W. of Wentworth, N.S.W.
Fig, 5—Two types of fossil ? burrows in ChowillaSand immediately under the BlanchetownClay at Bone Gulch (PL 6, fig. 2). The bur-rows are infilled with Blanchetown Clay.
Plate 8
Fig. 1—Dolomite Gulch, Triple Swamp, Moorna Sta-tion (Figs. 4, 22-24) W. of Wentworth,N.S.W.
GEOLOGY AND GEOMORPHOLOGY OF THE MURRAY RIVER 97
Fig. 2—Section of E. bank of Salt Creek, ScrubPaddock, Kulcuraa Stn. (Fig. 20), Cal Lai,N.S.W.
Fig. 3—Dolomitic limestone at Morkalla, NW. Vic-toria (Gray Bros, property), proof of lacus-trine conditions in what is now a semi-aridland. Ostracod fossils show the water wasfresh.
Fig. A—Close-up of dolomitic limestone of Fig. 3.
It is in situ and underlain by multi-colouredBlanchetown Clay.
Plate 9
Fig. 1—Triple Swamp from the mouth of DolomiteGulch, Moorna Station (Figs. 22-24), W. ofWentworth, N.S.W.
Fig. 2—W. side of Dolomite Gulch, Triple Swamp,Moorna Stn., showing the white bands ofdolomite. All the layers of dolomite knownin the study area are associated with the baseof the Blanchetown Clay.
Plate 10
Fig. 1—Mesa of Talgarry Sand, Nulla Stn., LakeVictoria lunette, N.S.W. The subhorizontalstratification should be noted. Unlike theunderlying Nulla Nulla Sand, the TalgarrySand has well-defined internal structures, con-tains only living species, and has a graycolumnar soil at the surface.
Fig. 2—Lake Victoria from the lunette on the E.side. The foreground shows Nulla Nulla Sand,and higher (behind the photographer) is theTalgarry Sand. The low slope near the lake(14° where measured) is Dunedin Park Sand,partly washed from the gulches and partlyblown up from the beach.
Fig. 3—Remains of an Aboriginal oven, consisting ofpieces of baked alluvium (used where thereare no stones) and charcoal. On sites of well-
sorted windblown sand, the oven stones some-times consist of relatively unsorted sand,showing that the material was carried up asmud from the river.
Plate 11
Fig. 1—Site of Aboriginal burials and middens in
Lybra Paddock, Keera Station, N. Victoria.The substrate is eroded Rufus Formation(Fig. 57). Site 16. The loose sand from ero-sion of the terrace is Dunedin Park Sand.
Fig. 2—Section in erosion gulch at Paringi where theSturt Highway touches the Murray R. tract
between Euston and Burnoga. BlanchetownClay overlies Parilla Sand, below which is
the mottled zone of a lateritic profile (Tim-beon Pedoderm).
—All photos by the author.
MEM. NAT. MUS. VICT. 34 PLATE 1
MEM. NAT. MUS. VICT. 34 PLATE 2
t -? in,
MEM. NAT. MUS. VICT. 34 PLATE 3
MEM. NAT. MUS. VICT. 34 PLATE 4
MEM. NAT. MUS. VICT. 34 PLATE 5
MEM. NAT. MUS. VICT. 34 PLATE 6
MEM. NAT. MUS. VICT. 34 PLATE 7
4\l^l -^WER$
VXrltr"
5*1
MI'IM. NA'l. MUS. VICT. 34 PLATE 8
MEM. NAT. MUS. VICT. 34 PLATE 9
V .,. ; znf,'-« ,
SiHifc.A
W#
MEM. NAT. MUS. VICT. 34 PLATE 10
*s*fm
MEM. NAT. MUS. VICT. 34 PLATE 11